Power grid equipment inspection method and device, computer equipment and storage medium

Abstract

The invention relates to a power grid equipment inspection method and device, computer equipment and a storage medium. The method comprises the following steps: according to power grid construction data and current position data of an unmanned aerial vehicle, constructing an inspection route of the unmanned aerial vehicle to power grid equipment; based on the inspection route and the size parameters of the unmanned aerial vehicle, controlling the unmanned aerial vehicle to carry out image acquisition on the to-be-detected equipment to be detected to obtain a target equipment image; and performing image analysis on the target equipment image according to a preset power grid equipment knowledge base to obtain a detection result of the to-be-detected equipment. Power grid construction data, unmanned aerial vehicle real-time perception and airborne intelligent analysis are fused, so that an unmanned aerial vehicle can autonomously plan an inspection route in a three-dimensional space, a safe channel is dynamically identified in a dense equipment network, an image is accurately acquired, and finally equipment state analysis and defect diagnosis are immediately completed based on an embedded knowledge base. The efficiency and the safety of inspection operation are improved, and the intelligent level and the active early warning capability of power grid operation and maintenance are enhanced.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a method, apparatus, computer equipment, and storage medium for inspecting power grid equipment. Background Technology

[0002] With the development of drone technology, the application scenarios for drones are becoming increasingly widespread, ranging from video shooting and cargo delivery to high-altitude inspection, with drones appearing in various industries. Power grid inspection is a crucial link in power grid safety, enabling the timely detection of potential hazards and ensuring the normal operation of the power grid.

[0003] In existing technologies, drones typically take photos of power grid equipment based on coordinates provided by personnel and then transmit the photos back for analysis. This method lacks the ability to actively identify power grid equipment, resulting in low inspection efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a method, device, computer equipment, and storage medium for inspecting power grid equipment that can improve inspection efficiency, in response to the above-mentioned technical problems.

[0005] Firstly, this application provides a method for inspecting power grid equipment, including:

[0006] Based on power grid construction data and the current location data of the drone, construct the inspection route of the drone for power grid equipment;

[0007] Based on the inspection route and the size parameters of the UAV, the UAV is controlled to acquire images of the device to be inspected, thereby obtaining images of the target device.

[0008] Image analysis is performed on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0009] In one embodiment, controlling the drone to capture images of the device under test based on the inspection route and the size parameters of the drone, and acquiring images of the target device, includes:

[0010] According to the inspection route, the drone is controlled to fly to the corresponding inspection range and acquire the target image data collected by the drone within the inspection range;

[0011] Based on the target image data and the size parameters of the drone, the passable space of the drone is determined;

[0012] Based on the size parameters, the inspection route, and the power grid construction data, the drone is controlled to fly and acquire images of the device under test.

[0013] In one embodiment, determining the passable space of the drone based on the target image data and the size parameters of the drone includes:

[0014] The volume frame of the UAV is constructed based on the size parameters of the UAV, and the passage frame of the UAV is generated based on the preset safety distance and the volume frame.

[0015] Based on the target image data, determine the size of the remaining space between each object within the inspection range;

[0016] The space between the objects is filtered based on the passage frame of the drone to determine the passage space of the drone.

[0017] In one embodiment, the power grid construction data includes distribution data of power grid equipment, spatial location data of power grid equipment, and equipment shape data of power grid equipment; the step of controlling the drone to fly to acquire images of the device under test based on the accessible size parameters, the inspection route, and the power grid construction data, and obtaining images of the target device, includes:

[0018] Based on the available dimensions and the distribution data of the power grid equipment, the inspection route is adjusted to obtain the target inspection route;

[0019] The power grid equipment is searched according to the target inspection route to determine the equipment to be tested.

[0020] Based on the spatial location and shape data of the device under test, an image of the device under test is acquired to obtain the target device image.

[0021] In one embodiment, the step of performing image analysis on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test includes:

[0022] Feature extraction is performed on the target device image to obtain the screen features of the device under test;

[0023] Anomaly identification is performed on the aforementioned image features to obtain abnormal device characteristics;

[0024] Based on the characteristics of the abnormal equipment, knowledge matching is performed in the power grid equipment knowledge base to obtain the target equipment knowledge;

[0025] Based on the knowledge of the target device, anomaly detection is performed on the abnormal device characteristics to obtain the detection results.

[0026] In one embodiment, constructing the inspection route of the UAV for power grid equipment based on power grid construction data and the current location data of the UAV includes:

[0027] The construction address is queried from the power grid construction data to obtain the construction addresses of towers and the distribution data of power grid equipment;

[0028] The tower construction addresses and power grid equipment distribution data are filtered according to the preset detection range to obtain the tower addresses and equipment distribution data to be tested.

[0029] The inspection range is determined based on the address of the tower to be tested and the distribution data of the equipment to be tested;

[0030] A starting point and an ending point are set at the boundary of the inspection range. Based on the starting point, the ending point, and the distribution data of the devices under test, the inspection route of the UAV for the power grid equipment is constructed.

[0031] In one embodiment, the method further includes:

[0032] The target device image and the detection result are encapsulated into a detection data packet;

[0033] The detection data packet is sent to the target terminal, and the validity of the detection data packet is evaluated.

[0034] Secondly, this application also provides an inspection device for power grid equipment, comprising:

[0035] The module is used to construct the inspection route of the UAV for the power grid equipment based on the power grid construction data and the current location data of the UAV;

[0036] The control module is used to control the UAV to acquire images of the device under test based on the inspection route and the size parameters of the UAV, thereby obtaining images of the target device.

[0037] The analysis module is used to perform image analysis on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0038] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0039] Based on power grid construction data and the current location data of the drone, construct the inspection route of the drone for power grid equipment;

[0040] Based on the inspection route and the size parameters of the UAV, the UAV is controlled to acquire images of the device to be inspected, thereby obtaining images of the target device.

[0041] Image analysis is performed on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0042] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0043] Based on power grid construction data and the current location data of the drone, construct the inspection route of the drone for power grid equipment;

[0044] Based on the inspection route and the size parameters of the UAV, the UAV is controlled to acquire images of the device to be inspected, thereby obtaining images of the target device.

[0045] Image analysis is performed on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0046] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0047] Based on power grid construction data and the current location data of the drone, construct the inspection route of the drone for power grid equipment;

[0048] Based on the inspection route and the size parameters of the UAV, the UAV is controlled to acquire images of the device to be inspected, thereby obtaining images of the target device.

[0049] Image analysis is performed on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0050] The aforementioned inspection methods, devices, computer equipment, and storage media for power grid equipment construct inspection routes for the UAVs based on power grid construction data and the current location data of the UAVs. Based on the inspection routes and the UAV's size parameters, the UAVs are controlled to acquire images of the equipment under test. Image analysis is then performed on the target equipment images using a pre-set power grid equipment knowledge base to obtain the inspection results. By integrating power grid construction data, real-time UAV perception, and onboard intelligent analysis, the UAVs can autonomously plan inspection routes in three-dimensional space, dynamically identify safe passages and accurately acquire images in dense equipment networks, and finally, perform real-time equipment status analysis and defect diagnosis based on the embedded knowledge base. This improves the efficiency and safety of inspection operations and enhances the intelligence level and proactive early warning capabilities of power grid operation and maintenance. Attached Figure Description

[0051] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a diagram illustrating the application environment of a power grid equipment inspection method in one embodiment.

[0053] Figure 2 This is a flowchart illustrating a method for inspecting power grid equipment in one embodiment;

[0054] Figure 3 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0055] Figure 4 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0056] Figure 5 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0057] Figure 6 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0058] Figure 7 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0059] Figure 8 This is a flowchart illustrating a method for inspecting power grid equipment in another embodiment;

[0060] Figure 9This is a structural block diagram of a power grid equipment inspection device in one embodiment;

[0061] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0063] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0064] The power grid equipment inspection method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, the target terminal 102 communicates with the drone 104 via a network. A data storage system can store the data that the drone 104 needs to process. The data storage system can be integrated onto the drone 104 or placed in the cloud or on other network servers. The drone 104 obtains power grid construction data from the target terminal, determines the inspection route based on the power grid construction data, performs inspection operations, and sends the inspection results to the target terminal 102. The target terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart vehicle devices, projection devices, etc. Portable wearable devices can include smartwatches, smart bracelets, head-mounted devices, etc. Head-mounted devices can be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc.

[0065] In one exemplary embodiment, such as Figure 2 As shown, a method for inspecting power grid equipment is provided, which can be applied to... Figure 1 The following explanation uses the inspection of power grid equipment as an example, including:

[0066] S201: Based on power grid construction data and the current location data of the UAV, construct the inspection route of the UAV for power grid equipment.

[0067] The power grid construction data includes the geographical coordinates and spatial distribution information of power equipment within the power grid. This equipment can include poles, insulator strings, surge arresters, and more.

[0068] In this embodiment, the UAV acquires power grid construction data within the mission area and simultaneously obtains its own location data in real time through its onboard positioning system. This location data includes latitude and longitude, altitude, pitch angle, roll angle, and heading angle.

[0069] In this embodiment, based on the received power grid construction data and real-time location data, inspection route planning is performed. The current 3D coordinates of the UAV can be used as the starting point of the path, transforming all the equipment points to be tested in the power grid construction data into necessary 3D waypoints. Further, a globally optimal waypoint access sequence is calculated in 3D space, aiming to minimize the total flight distance and time. Optionally, after generating the basic sequence, the algorithm further performs spatial smoothing processing, performs curve fitting and elevation optimization on the point-to-point path based on the UAV's dynamic performance, and automatically avoids known geographical no-go zones and airspace restrictions to generate inspection routes for power equipment.

[0070] S202, based on the inspection route and the size parameters of the UAV, controls the UAV to acquire images of the equipment to be inspected.

[0071] In this embodiment, the UAV begins flight according to the inspection route. When it arrives within the preset observation airspace of a device under test, it first precisely retrieves all its dimensional parameters from its local database, such as the maximum fuselage outline size, the diameter of the dynamic envelope circle defined by the rotor rotation, and the gimbal extension at extreme angles. Simultaneously, the UAV uses an image acquisition system and LiDAR to synchronously scan the device under test and its surrounding environment, generating dense depth point clouds and high-resolution images in real time, and fusing them into target image data with depth information. Furthermore, the spatial relationship between the safety boundary model defined by the dimensional parameters and the three-dimensional point cloud model of the real-time environment is analyzed. Through a preset path search algorithm, one or more continuous three-dimensional spatial regions that can completely accommodate the UAV's safety boundary and have no collision risk are determined, i.e., dynamically passable spaces.

[0072] Furthermore, after confirming that the space is passable, a flight path is planned to enter the safe space from the current position, and the aircraft maneuvers to the selected optimal observation point for stable hovering. Then, the gimbal is adjusted to align with the device under test with the preset optimal pitch and yaw angles, and the focus and exposure parameters are automatically adjusted to complete the image acquisition of the target device and obtain the image of the target device.

[0073] S203, perform image analysis on the target equipment image based on the preset power grid equipment knowledge base to obtain the detection result of the equipment under test.

[0074] Optionally, the power grid equipment knowledge base can be an indicator library that integrates equipment digital models, historical defect feature maps, and deep learning diagnostic models.

[0075] In this embodiment, the acquired raw target device image is first preprocessed, including distortion correction, illumination equalization, and region of interest (ROI) segmentation. Furthermore, multi-layered visual features can be extracted from the preprocessed image, including surface texture statistics, edge gradient histograms, color distribution, and specific shape contours.

[0076] Optionally, the feature vector can be input into a diagnostic neural network model in the knowledge base. The model compares and calculates the probability of the input features with the normal state baseline and defect pattern features in the database in a high-dimensional space to identify whether there are abnormalities such as cracks, heat spots or structural deformation in the image, and forms a complete detection record. The detection result can be output in a standardized data format. The detection result can include the device ID, health status classification, specific defect type and code, severity level assessment, pixel-level location box coordinates of the defect in the image, and analysis confidence.

[0077] In the aforementioned power grid equipment inspection method, an inspection route for the power grid equipment is constructed based on power grid construction data and the current location data of the UAV. Based on the inspection route and the UAV's size parameters, the UAV is controlled to acquire images of the equipment under test. Image analysis is then performed on the target equipment images using a pre-set power grid equipment knowledge base to obtain the inspection results. By integrating power grid construction data, real-time UAV perception, and onboard intelligent analysis, the UAV can autonomously plan inspection routes in three-dimensional space, dynamically identify safe passages and accurately acquire images in dense equipment networks, and finally, perform real-time equipment status analysis and defect diagnosis based on the embedded knowledge base. This improves the efficiency and safety of inspection operations and enhances the intelligence level and proactive early warning capabilities of power grid operation and maintenance.

[0078] In one embodiment, one implementation of S203 above is provided, such as... Figure 3 As shown, the above-mentioned "performing image analysis on the target equipment image based on a preset power grid equipment knowledge base to obtain the detection result of the equipment under test" includes:

[0079] S301 controls the drone to fly to the corresponding inspection range according to the inspection route and acquires the target image data collected by the drone within the inspection range.

[0080] In this embodiment, the UAV flies to the designated inspection airspace for the towers according to the generated inspection route. Upon entering the inspection area, the UAV utilizes its onboard image acquisition equipment and LiDAR. The image acquisition equipment continuously and at high resolution captures images of key components such as the tower body, insulator strings, hardware, and conductor connection points along a preset scanning path. The LiDAR performs a high-speed rotating scan of the same space to acquire 3D point cloud data of the environment. The multimodal sensor data is then fused in real time, precisely associating 2D image pixels with 3D spatial points to generate a target image dataset integrating visual and depth information and bearing spatial coordinate labels.

[0081] For example, taking a power grid inspection task in the eastern district of a city as an example, the drone departs from the starting point A according to the initial flight path, flies 8 kilometers along the designated route, and enters the inspection area. Within the inspection area, the drone activates its camera to collect surrounding images, capturing a total of 120 high-definition images. Distance recognition is performed on each image, identifying the nearest tower as 50 meters away and the farthest tree as 200 meters away. A preset distance threshold of 150 meters is set. When a tree is identified as being 200 meters away (exceeding the threshold), the image quality of that area is reduced from 4K to 720P, generating a low-quality area. Subsequently, the low-quality area is cropped, retaining the central area (occupying 60% of the original image) to obtain the high-quality target image area. Object feature recognition is performed on the target image area, identifying the metal support features of the tower (cylindrical shape, 30 meters high), the insulator string features (white porcelain, 2 meters long), and the surrounding environment features (trees 10 meters high, grass coverage 80%). Based on these characteristics, the target image area data is digitized into structured information, including tower coordinates (116.46°E, 39.91°N), equipment type (transformer), list of environmental obstacles (trees, utility poles), etc., and finally a target image data package is generated.

[0082] S302 determines the space through which the drone can pass, based on target image data and the drone's size parameters.

[0083] In this embodiment, after completing environmental data acquisition and generating target image data, the UAV initiates real-time obstacle avoidance and passage planning calculations. First, a refined 3D environment model constructed from LiDAR point clouds is parsed from the target image data. Then, a model representing the UAV's own dimensions is used to perform real-time collision detection simulations with the 3D environment model. All spatial areas in the environment that can accommodate the UAV without collision risk are identified. These continuous 3D spaces that conform to the UAV's maneuverability are defined as passable spaces under the current conditions.

[0084] The S303, based on available size parameters, inspection routes, and power grid construction data, controls the flight of a drone to acquire images of the equipment under test.

[0085] In this embodiment, within the boundary of the passable space, the trajectory near the current target point in the inspection route is locally replanned to generate a new local route entirely within the safe passage. When approaching the device under test, the UAV, based on the type of the device, retrieves its standard 3D model and optimal shooting angle requirements from the power grid construction data. This allows it to search within the current local passable space and calculate an optimal observation point, ensuring the UAV is in a safe position and that the camera is aligned with the device's surface at the specified angle and distance. After determining the target point, the UAV autonomously controls its propulsion system to fly along the safe path to that point, precisely adjusting its flight attitude and gimbal orientation to acquire images of the device under test, thus obtaining an image of the target device.

[0086] In the above-mentioned embodiments, the UAV autonomously constructs an accurate passable space model based on real-time collected environmental data, and uses this as an absolute safety boundary to dynamically guide its flight and shooting missions, ensuring the safety of the UAV in close-range operations in complex and dense power grids and avoiding collision risks. By performing real-time fusion calculations of the safety space, mission path and equipment model, the UAV can obtain high-definition and standardized diagnostic images from the optimal perspective.

[0087] In one embodiment, one implementation of the above-described S302 is provided, such as... Figure 4 As shown, the above-mentioned "determining the passable space of the drone based on target image data and the drone's size parameters" includes:

[0088] S401 constructs the volume frame of the UAV based on its size parameters, and generates the passage frame of the UAV based on the preset safety distance and volume frame.

[0089] In the embodiment of this application, the UAV obtains its complete dimensional parameters locally, including the length, width, and height of the main fuselage, as well as the length of each rotor arm and the position of its motor. Based on these dimensional parameters, a minimal enclosing geometry, i.e., a volumetric frame, is constructed that fits the physical contours of the fuselage. Further, according to preset flight safety procedures, the UAV invokes a multi-dimensional safety distance parameter set, uniformly extending the safety distance outward to every surface, edge, and corner of the volumetric frame. Through three-dimensional offset calculations, a new, larger, and smoother virtual geometry, i.e., the frame itself, is generated.

[0090] Optionally, the set of safety distance parameters includes static minimum clearance requirements between the aircraft and obstacles, as well as rotor airflow disturbance range, sensor measurement errors, and position estimation uncertainties during dynamic flight.

[0091] S402, based on the target image data, determine the size of the space between each object within the inspection range.

[0092] In this embodiment, visual and lidar information from the target image data are fused to generate a dense point cloud with three-dimensional coordinates and object category labels. Based on the point cloud's reflection intensity, continuity, and normal vector characteristics, point clouds representing different objects such as conductors, insulators, tower structures, and trees are clustered and segmented to identify individual objects within the inspection range. Furthermore, spatial relationship analysis is performed on any two or more spatially adjacent or intersecting objects. By calculating the minimum distance between these object point clouds and combining this with the object's geometry, the actual size of the unoccupied three-dimensional gaps between objects—that is, the reserved space size—is determined.

[0093] S403, based on the drone's pass-through frame, filters the space between objects to determine the drone's pass-through space.

[0094] In this embodiment of the application, the flight mode of the drone is obtained, the variable angle of the drone is obtained based on the flight mode, and the extreme passage mode of the drone is constructed based on the variable angle and the frame. The remaining space is filtered by combining the frame and the extreme passage mode to obtain the passable space around the drone.

[0095] Optionally, the pass-through frame, representing the required safe space, is used as a rigid 3D template and simulated to match the size of each available space calculated in the environment. The pass-through frame model is then attempted to be placed in the 3D region corresponding to a specific available space, checking whether it is completely contained within that space and has no intersection with the point cloud of objects constituting the space's boundary, in order to identify feasible passage methods. All available spaces that can completely accommodate the pass-through frame (with at least one feasible pose) are marked as passable spaces.

[0096] For example, taking a certain type of power drone as an example, its dimensions are 1.2 meters long, 0.8 meters wide, and 0.5 meters high. A cuboid frame (1.2 meters long × 0.8 meters wide × 0.5 meters high) is constructed based on these parameters, with a safety distance of 0.5 meters. After expansion, a passage frame (2.2 meters long × 1.8 meters wide × 1.5 meters high) is generated. Based on the target image data, the intersecting images of objects around a certain tower in the eastern area are extracted, identifying the intersecting relationship between the power transmission lines (10 meters apart) between two adjacent towers and the trees below (15 meters high). Through spatial relationship analysis, the vertical distance between the power transmission lines and the trees is calculated to be 5 meters, the horizontal distance to be 3 meters, and the remaining space size is 8 meters long × 4 meters wide × 5 meters high. The drone flies in multi-rotor mode with variable angles of ±30° pitch and ±45° roll. Combining the passage frame and the variable angles, the calculated ultimate passage method is a 30° tilt side-fly. After screening the available space, it was determined that the area with a length of 6 meters, a width of 3 meters, and a height of 4 meters could be used to cover the safety passage under the power transmission line.

[0097] In the above-mentioned application embodiments, safe flight is upgraded from relying on experience-based judgment or reactive obstacle avoidance to deterministic planning based on accurate models and forward-looking calculations, which ensures the feasibility and safety of every action of the UAV in the inspection environment with intersecting wires and dense structures.

[0098] In one embodiment, one implementation of the above-described S303 is provided, wherein the power grid construction data includes distribution data of power grid equipment, spatial location of power grid equipment, and equipment shape data of power grid equipment; such as Figure 5 As shown, the aforementioned "controlling the drone to acquire images of the equipment under test based on available size parameters, inspection routes, and power grid construction data" includes:

[0099] S501, based on the available dimensions and distribution data of power grid equipment, adjusts the inspection route to obtain the target inspection route.

[0100] In this embodiment, after obtaining the set of accessible spaces, the UAV finely adjusts the inspection route based on the accessible spaces and the distribution data of the devices to be tested, thus obtaining the target inspection route. Optionally, the three-dimensional boundary of the accessible spaces can be used as a rigid safety fence for flight, and the distribution data of power grid equipment in the power grid construction data can be used as the set of target points that must be visited.

[0101] For example, the system can first check whether each segment and waypoint in the original inspection route is completely within the current passable space. For any segment that intersects with or is outside the safety space boundary, the system will use the corresponding passable space as the feasible region and employ a path search algorithm to recalculate an alternative local path that is completely within the safety space. Simultaneously, based on the spatial relationship of the device distribution, and while ensuring the overall path is coherent and efficient, the local order of device access is fine-tuned to ultimately generate a target inspection route that matches the real-time environmental safety conditions.

[0102] S502 searches for power grid equipment according to the target inspection route and determines the equipment to be tested.

[0103] In this embodiment, a saturation search is performed on the power grid equipment according to the target inspection route, and multiple screenings are conducted to obtain the target power grid equipment to be inspected. Optionally, while the UAV flies along the route, the onboard visual recognition system continuously analyzes the environment in front and to the sides in real time, fusing LiDAR point clouds with visible light images to quickly identify and classify power equipment objects in the scene. When a device object is identified, a multiple screening process is initiated, power grid construction data is retrieved, and the visual features and spatial location of the device are compared with the data archive to confirm whether it is a device within the current mission plan and to verify its identity. Devices that are successfully identified, have matched identities, and have good detection conditions are determined as the devices to be tested.

[0104] S503 acquires images of the device under test based on the spatial location and shape data of the device under test.

[0105] In this embodiment of the application, the current spatial location, shape data of the target power grid device, and images of surrounding obstacles are obtained. Based on the current spatial location and shape data, multiple optimal shooting points are obtained. Based on the images of surrounding obstacles, the multiple optimal shooting points are filtered to obtain the target shooting point. The relative spatial angle between the target shooting point and the target power grid device is extracted. The shooting angle of the drone is adjusted according to the relative spatial angle before shooting to obtain an image of the target device.

[0106] In the above-mentioned embodiments, the UAV can dynamically reconstruct an adjusted inspection route that is entirely within the passable space and efficiently covers all target devices. The UAV performs active search and intelligent screening based on real-time perception along this safe route, which improves the accuracy of target recognition and adaptability to complex on-site situations, thereby stably and reliably acquiring images of target devices.

[0107] In one embodiment, one implementation of S203 above is provided, such as... Figure 6As shown, the above-mentioned "performing image analysis on the target equipment image based on a preset power grid equipment knowledge base to obtain the detection result of the equipment under test" includes:

[0108] S601, extract features from the target device image to obtain the screen features of the device under test.

[0109] In this embodiment, after acquiring an image of the target device, the UAV performs feature extraction on the image. Optionally, the image can be preprocessed, including noise reduction, contrast enhancement, and color space conversion. Furthermore, multiple feature extraction algorithms can be run in parallel. For example, the microscopic texture features of the device surface can be quantified using histogram of oriented gradients or local binary mode algorithms; the geometric shape and structural features of the device and its components can be obtained through edge detection and contour extraction algorithms; and the color histogram of the image and the statistical distribution of pixel values ​​in specific areas can be analyzed to obtain color and brightness features. This results in a multi-dimensional set of image features.

[0110] S602 performs anomaly identification on screen features to obtain abnormal device characteristics.

[0111] In this embodiment, after obtaining the image features, anomaly identification is performed on these features. Optionally, one or more lightweight, pre-trained deep learning models or traditional machine learning classifiers can be used for anomaly identification. Using the extracted image features as input, the model performs layer-by-layer calculations and feature mapping, comparing the input image features with an embedded benchmark. When feature values ​​deviate from the expected distribution, the model identifies and marks these deviations. For example, a sudden drop in the texture uniformity index of the insulator skirt area, or an abnormal shift in the color channel values ​​at the metal connection. Further, all marked deviation features are summarized and correlated with their corresponding pixel coordinates to output a set of abnormal device feature data.

[0112] S603: Based on the characteristics of the abnormal equipment, perform knowledge matching in the power grid equipment knowledge base to obtain the knowledge of the target equipment.

[0113] In this embodiment, abnormal equipment characteristics are used as query conditions for multi-dimensional matching and retrieval in the knowledge base. Specifically, known defects with similar abnormal characteristic patterns can be searched in the fault mode library of the knowledge base; the final diagnostic conclusions of the same model of equipment exhibiting similar characteristics under similar environments and operating conditions in the case library of the knowledge base can be retrieved. Thus, the fault mechanism, typical defect images, and development patterns most relevant to the abnormal equipment characteristics are identified from the knowledge base, which constitutes the target equipment knowledge.

[0114] Optionally, multiple historical detection data are acquired, and the process of the multiple historical detection data is decomposed to obtain multiple detection processes and multiple detection results; the multiple detection processes and multiple detection results are correlated and mutually verified to generate a historical detection neural network; equipment knowledge is extracted from the historical detection neural network to generate a neural network knowledge set, and a power grid equipment knowledge base is constructed based on the neural network knowledge set.

[0115] S604, based on the knowledge of the target equipment, perform anomaly detection on the characteristics of abnormal equipment and obtain the detection results.

[0116] In this embodiment, a final diagnostic decision is made based on the target device knowledge obtained through matching. According to rules or discrimination models in the knowledge base, the combination, intensity, and spatial distribution of abnormal device features are comprehensively evaluated to determine which specific defect mode it conforms to. For example, the knowledge might indicate the feature of "abnormal local texture roughness of the insulator skirt." If accompanied by a slight color shift in a specific area, the probability of it pointing to "early stage of electrical erosion" is as high as 85%; if it is an isolated feature, it may simply be "contamination deposition." Further, a structured detection result is generated. Optionally, the detection result may include the defect type, the precise location of the defect, the severity level, the diagnostic confidence level, and a preliminary maintenance recommendation.

[0117] In the above-mentioned embodiments, the accuracy and efficiency of defect identification are improved, transforming the traditional lagging judgment that relies on human experience into real-time automatic diagnosis, and the diagnostic capabilities of the system are continuously optimized through continuous learning of the knowledge base.

[0118] In one embodiment, one implementation of S201 above is provided, such as... Figure 7 As shown, the above-mentioned "constructing an inspection route for power grid equipment by drone based on power grid construction data and the current location data of the drone" includes:

[0119] S701 performs a construction address query on the power grid construction data to obtain the construction addresses of towers and the distribution data of power grid equipment.

[0120] In this embodiment, the UAV receives power grid construction data from a target terminal. The UAV performs in-depth querying and parsing of the power grid construction data. Optionally, data extraction is performed on the geographic information field and equipment attribute field in the power grid construction data to obtain the construction location information of all poles and the model, relative mounting position, and status identifier of various equipment on these poles, forming structured equipment distribution data. The construction location information is a set of tower construction addresses including latitude, longitude, and altitude.

[0121] S702, based on the preset detection range, filters the tower construction address and power grid equipment distribution data to obtain the tower address and equipment distribution data to be tested.

[0122] In this embodiment, spatial filtering is performed based on a preset detection range issued by the control center. The spatial location of each tower construction address is calculated relative to the preset detection range, and all addresses falling within the preset detection range are filtered out to form a list of tower addresses to be tested. Simultaneously, based on the affiliation between equipment distribution data and the aforementioned tower addresses, all equipment information carried on these towers to be tested is simultaneously filtered out, generating distribution data of the equipment to be tested.

[0123] S703 determines the inspection range based on the address of the tower to be tested and the distribution data of the equipment to be tested.

[0124] In this embodiment, based on the filtered list of tower addresses to be tested and their corresponding distribution data of the equipment to be tested, the UAV performs aggregate calculations of the inspection space. It does not simply use a preset detection range, but rather calculates a minimum three-dimensional spatial region that can tightly enclose all tower addresses to be tested, based on the actual geographical clusters of the targets. This region fully considers the height distribution of the equipment and the space margin required for safe operation, and is ultimately determined as the specific and precise inspection range for this task. This range in three-dimensional space may be more accurate and efficient than the original two-dimensional preset range.

[0125] S704 sets the start and end points at the boundaries of the inspection range, and constructs the inspection route for the power grid equipment by drone based on the start and end points and the distribution data of the equipment to be tested.

[0126] In this embodiment, two points are selected on the spatial boundary of the calculated inspection range: one as the starting point of the task (usually close to the current location of the UAV or an easily accessible location), and the other as the ending point of the task (usually an easily accessible location for return or relocation after the inspection). Further, the UAV uses the starting point as the beginning of its path and the ending point as the end point, treating all the device points to be inspected in the distribution data of the devices under test as necessary waypoints. The UAV employs a path planning algorithm to calculate the optimal sequence and path connecting all device points and linking the starting and ending points within the inspection range, aiming for the shortest total flight distance or the optimal time, ultimately constructing a complete inspection route.

[0127] For example, taking the inspection task of a city's power grid system as an example, the power grid construction data includes the location information of towers, substation coordinates, and transmission line layout within the area. First, the construction address is queried from the power grid construction data, revealing a total of 50 towers distributed across two main areas, east and west. Equipment distribution data shows 15 transformers and 20 circuit breakers. Based on the approximate inspection range (entire eastern area and northern western area), 30 towers in the eastern area and 10 towers in the northern western area are selected as the towers to be inspected. Simultaneously, 12 transformers in the eastern area and 8 circuit breakers in the northern western area are selected as the equipment distribution data to be inspected. Combining the tower addresses and equipment distribution data, the inspection range is determined to be approximately 5 square kilometers, encompassing the entire eastern area and the northern western area. The starting point is set at the eastern boundary point A (coordinates: 116.45°E, 39.90°N), and the ending point is set at the northern western boundary point B (coordinates: 116.40°E, 39.88°N). Based on the starting point A and the current position of the UAV (coordinates: 116.50°E, 39.95°N), the positioning route is generated as a straight path with a length of approximately 8 kilometers. Based on the starting point A, the ending point B, and the distribution data of the equipment to be tested, the planned inspection route is an S-shaped path that circles the tower in the eastern area and connects to the northern part of the western area, with a length of approximately 25 kilometers. Combining the positioning route and the inspection route, an initial route with a total length of approximately 33 kilometers is generated.

[0128] In the above-mentioned application embodiments, by intelligently setting the start and end points and comprehensively optimizing the global path by integrating all test points, a closed and efficient inspection route is constructed, which shortens the task preparation time, reduces the risk of human error, and ensures the economy and rationality of the flight path itself, laying the foundation for subsequent autonomous, safe and efficient inspection operations.

[0129] In one embodiment, such as Figure 8 As shown, the above-mentioned inspection method for power grid equipment also includes:

[0130] S801 encapsulates the target device image and detection results into a detection data packet.

[0131] S802 sends the detection data packet to the target terminal and evaluates the validity of the detection data packet.

[0132] In this embodiment, the data transmission location of the UAV is obtained, the target device image and detection results are organized to generate a detection data packet, key data in the detection data packet is extracted, and an initial data stream is generated. A return switch is added to the initial data stream to generate a return data stream, which the UAV transmits back to the target terminal. After receiving the return data stream, the target terminal triggers the return switch, which transmits the initial data stream back to the UAV. After receiving the initial data stream, the UAV extracts the complete value of the initial data stream and performs a transmission security assessment on the data transmission location based on the complete value to obtain a transmission security value. It is determined whether the transmission security value is higher than a preset security threshold. If the transmission security value is determined to be lower than the security threshold, the current surrounding image of the UAV is obtained. The data transmission location of the UAV is replanned based on the current surrounding image to obtain a target transmission location with better data transmission effect, and the detection data packet is transmitted back to the target terminal at the target transmission location.

[0133] Optionally, a data black box can be installed on the drone to collect the target device images and detection results, generate a detection data package, transmit the detection data package back to the staff, generate a backup data package and store it in the data black box, and evaluate the validity of the detection data package to obtain the valid data value. Based on the valid data value, the data black box can be retrieved.

[0134] Optionally, the process of evaluating the validity of the detection data packet may include: evaluating the validity of the detection data packet to obtain the valid data value of the detection data packet; determining whether the valid data value is higher than a preset valid threshold or whether the UAV has received a return-to-home command; if the valid value is not higher than the valid threshold or the UAV has received a return-to-home command, the UAV obtains its current position and the position of the target terminal; based on the current position of the UAV and the position of the target terminal, planning and generating multiple return-to-home routes and obtaining the return-to-home time for each route; selecting the return-to-home route with the shortest return-to-home time as the target return-to-home route, and the UAV returns according to the target return-to-home route to retrieve the data black box.

[0135] In one embodiment, a complete method for inspecting power grid equipment is provided, including:

[0136] S1 performs a construction address query on the power grid construction data to obtain the construction addresses of towers and the distribution data of power grid equipment.

[0137] S2, based on the preset detection range, filter the tower construction address and power grid equipment distribution data to obtain the tower address and equipment distribution data to be tested.

[0138] S3. Determine the inspection range based on the address of the tower to be tested and the distribution data of the equipment to be tested.

[0139] S4 sets the starting and ending points at the boundaries of the inspection range, and constructs the inspection route for the power grid equipment by the drone based on the starting and ending points and the distribution data of the equipment to be tested.

[0140] S5 controls the drone to fly to the corresponding inspection range according to the inspection route, and acquires the target image data collected by the drone within the inspection range.

[0141] S6: Construct the volume frame of the drone based on the drone's size parameters, and generate the drone's passage frame based on the preset safety distance and volume frame.

[0142] S7 determines the size of the space between objects within the inspection range based on the target image data.

[0143] S8 filters the space between objects based on the drone's pass-through frame to determine the drone's pass-through space.

[0144] S9 adjusts the inspection route based on the available dimensions and distribution data of power grid equipment to obtain the target inspection route.

[0145] S10: Search for power grid equipment according to the target inspection route and determine the equipment to be tested.

[0146] S11. Based on the spatial location and shape data of the device under test, image acquisition is performed on the device under test to obtain the target device image.

[0147] S12, extract features from the target device image to obtain the screen features of the device under test.

[0148] S13, perform anomaly identification on screen features to obtain abnormal device features.

[0149] S14. Based on the characteristics of the abnormal equipment, perform knowledge matching in the power grid equipment knowledge base to obtain the knowledge of the target equipment.

[0150] S15, based on the knowledge of the target equipment, perform anomaly detection on the abnormal equipment characteristics and obtain the detection results.

[0151] S16, encapsulate the target device image and detection results into a detection data packet.

[0152] S17. Send the detection data packet to the target terminal and evaluate the validity of the detection data packet.

[0153] In the aforementioned power grid equipment inspection method, an inspection route for the power grid equipment is constructed based on power grid construction data and the current location data of the UAV. Based on the inspection route and the UAV's size parameters, the UAV is controlled to acquire images of the equipment under test. Image analysis is then performed on the target equipment images using a pre-set power grid equipment knowledge base to obtain the inspection results. By integrating power grid construction data, real-time UAV perception, and onboard intelligent analysis, the UAV can autonomously plan inspection routes in three-dimensional space, dynamically identify safe passages and accurately acquire images in dense equipment networks, and finally, perform real-time equipment status analysis and defect diagnosis based on the embedded knowledge base. This improves the efficiency and safety of inspection operations and enhances the intelligence level and proactive early warning capabilities of power grid operation and maintenance.

[0154] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0155] Based on the same inventive concept, this application also provides a power grid equipment inspection device for implementing the above-described power grid equipment inspection method. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more power grid equipment inspection device embodiments provided below can be found in the limitations of the power grid equipment inspection method described above, and will not be repeated here.

[0156] In one embodiment, such as Figure 9 As shown, a power grid equipment inspection device is provided, comprising: a construction module 10, a control module 11, and an analysis module 12, wherein:

[0157] The construction module 10 is used to construct the inspection route of the UAV for the power grid equipment based on the power grid construction data and the current location data of the UAV.

[0158] Control module 11 is used to control the UAV to acquire images of the device under test based on the inspection route and the size parameters of the UAV, and to obtain images of the target device.

[0159] The analysis module 12 is used to perform image analysis on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

[0160] In one embodiment, the control module 11 includes: a flight unit, a determination unit, and a data acquisition unit, wherein:

[0161] The flight unit is used to control the UAV to fly to the corresponding inspection range according to the inspection route, and to acquire the target image data collected by the UAV within the inspection range.

[0162] The determining unit is used to determine the passable space of the drone based on the target image data and the size parameters of the drone.

[0163] The acquisition unit is used to control the UAV to fly to acquire images of the device under test based on the available size parameters, the inspection route, and the power grid construction data, thereby obtaining images of the target device.

[0164] In one embodiment, the aforementioned determining unit is specifically configured to construct the volume frame of the UAV based on the UAV's size parameters, and generate the UAV's passage frame based on a preset safety distance and the volume frame; determine the size of the space between each object within the inspection range based on the target image data; and filter the size of the space between each object based on the UAV's passage frame to determine the UAV's passable space.

[0165] In one embodiment, the aforementioned acquisition unit is specifically used to adjust the inspection route according to the passable size and the distribution data of the power grid equipment to obtain a target inspection route; search for power grid equipment according to the target inspection route to determine the equipment to be tested; and acquire an image of the equipment to be tested according to the spatial location and shape data of the equipment to be tested to obtain an image of the target equipment.

[0166] In one embodiment, the analysis module 12 includes: an extraction unit, an identification unit, a matching unit, and an analysis unit, wherein:

[0167] The extraction unit is used to extract features from the target device image to obtain the screen features of the device under test.

[0168] The identification unit is used to identify anomalies in the screen features and obtain abnormal device features.

[0169] The matching unit is used to perform knowledge matching in the power grid equipment knowledge base based on the characteristics of the abnormal equipment to obtain the target equipment knowledge.

[0170] The analysis unit is used to perform anomaly detection on the abnormal device characteristics based on the target device knowledge, and obtain the detection result.

[0171] In one embodiment, the construction module 10 includes: a query unit, a filtering unit, a range determination unit, and a construction unit, wherein:

[0172] The query unit is used to query the construction address of the power grid construction data to obtain the construction address of the tower and the distribution data of the power grid equipment.

[0173] The filtering unit is used to filter the tower construction address and the power grid equipment distribution data according to a preset detection range to obtain the tower address and equipment distribution data to be tested.

[0174] The range determination unit is used to determine the inspection range based on the address of the tower to be tested and the distribution data of the equipment to be tested.

[0175] The construction unit is used to set the starting position and the ending position at the boundary of the inspection range, and to construct the inspection route of the UAV for the power grid equipment based on the starting position, the ending position and the distribution data of the equipment under test.

[0176] In one embodiment, the above-mentioned power grid equipment inspection device further includes: an encapsulation module and a transmission module, wherein:

[0177] An encapsulation module is used to encapsulate the target device image and the detection result into a detection data packet.

[0178] The sending module is used to send the detection data packet to the target terminal and to evaluate the validity of the detection data packet.

[0179] Each module in the aforementioned power grid equipment inspection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0180] In one embodiment, a computer device is provided, which may be a drone, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for inspecting power grid equipment. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0181] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0182] In one embodiment, a computer device is provided having a computer program stored thereon, which, when executed by a processor, implements the various processes in the above method embodiments.

[0183] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the various processes in the above method embodiments.

[0184] This application also provides a computer program product containing instructions that, when run on a computer, implements the various processes in the above method embodiments.

[0185] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0186] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0187] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for inspecting power grid equipment, characterized in that, The method includes: Based on power grid construction data and the current location data of the drone, construct the inspection route of the drone for power grid equipment; Based on the inspection route and the size parameters of the UAV, the UAV is controlled to acquire images of the device to be inspected, thereby obtaining images of the target device. Image analysis is performed on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

2. The method according to claim 1, characterized in that, The step of controlling the drone to capture images of the device under test based on the inspection route and the size parameters of the drone, and acquiring images of the target device, includes: According to the inspection route, the drone is controlled to fly to the corresponding inspection range and acquire the target image data collected by the drone within the inspection range; Based on the target image data and the size parameters of the drone, the passable space of the drone is determined; Based on the size parameters, the inspection route, and the power grid construction data, the drone is controlled to fly and acquire images of the device under test.

3. The method according to claim 2, characterized in that, The step of determining the passable space of the drone based on the target image data and the size parameters of the drone includes: The volume frame of the UAV is constructed based on the size parameters of the UAV, and the passage frame of the UAV is generated based on the preset safety distance and the volume frame. Based on the target image data, determine the size of the remaining space between each object within the inspection range; The space between the objects is filtered based on the passage frame of the drone to determine the passage space of the drone.

4. The method according to claim 2, characterized in that, The power grid construction data includes distribution data of power grid equipment, spatial location data of power grid equipment, and equipment shape data of power grid equipment; the step of controlling the drone to fly to acquire images of the equipment under test based on the accessible size parameters, the inspection route, and the power grid construction data, and obtaining images of the target equipment, includes: Based on the available dimensions and the distribution data of the power grid equipment, the inspection route is adjusted to obtain the target inspection route; The power grid equipment is searched according to the target inspection route to determine the equipment to be tested. Based on the spatial location and shape data of the device under test, an image of the device under test is acquired to obtain the target device image.

5. The method according to claim 1, characterized in that, The step of performing image analysis on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test includes: Feature extraction is performed on the target device image to obtain the screen features of the device under test; Anomaly identification is performed on the aforementioned image features to obtain abnormal device characteristics; Based on the characteristics of the abnormal equipment, knowledge matching is performed in the power grid equipment knowledge base to obtain the target equipment knowledge; Based on the knowledge of the target device, anomaly detection is performed on the abnormal device characteristics to obtain the detection results.

6. The method according to claim 1, characterized in that, The step of constructing the inspection route for power grid equipment by the drone based on power grid construction data and the drone's current location data includes: The construction address is queried from the power grid construction data to obtain the construction addresses of towers and the distribution data of power grid equipment; The tower construction addresses and power grid equipment distribution data are filtered according to the preset detection range to obtain the tower addresses and equipment distribution data to be tested. The inspection range is determined based on the address of the tower to be tested and the distribution data of the equipment to be tested; A starting point and an ending point are set at the boundary of the inspection range. Based on the starting point, the ending point, and the distribution data of the devices under test, the inspection route of the UAV for the power grid equipment is constructed.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: The target device image and the detection result are encapsulated into a detection data packet; The detection data packet is sent to the target terminal, and the validity of the detection data packet is evaluated.

8. An inspection device for power grid equipment, characterized in that, The device includes: The module is used to construct the inspection route of the UAV for the power grid equipment based on the power grid construction data and the current location data of the UAV; The control module is used to control the UAV to acquire images of the device under test based on the inspection route and the size parameters of the UAV, and to obtain images of the target device. The analysis module is used to perform image analysis on the target device image based on a preset power grid equipment knowledge base to obtain the detection result of the device under test.

9. 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, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images