Power transmission line adaptive inspection method and system

Abstract

The invention provides a power transmission line adaptive inspection method and system, and belongs to the technical field of power transmission inspection. According to the real-time image, a light-weight visual navigation model is adopted to determine and recognize a tower, and the unmanned aerial vehicle flies to the tower top of the tower in an imitated mode according to the tower shape features of the recognized tower; wherein in the simulated flight, a GPS and a visual SLAM are adopted to independently solve a pose, and an optimally estimated fusion height is output through extended Kalman filtering fusion; adopting an identification algorithm for the tower to obtain tower parameters; and determining a routing inspection waypoint plan according to the tower parameter matching database, and shooting a routing inspection image of the tower key equipment according to the routing inspection waypoint plan. According to the invention, algorithms of target identification, visual tracking, automatic obstacle avoidance, path correction, adaptive dimming, automatic focusing, multi-angle shooting and the like are integrated, and power transmission line inspection with low difficulty, high efficiency and accurate identification is realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of power transmission inspection, and in particular relates to a method and system for adaptive inspection of power transmission lines. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Transmission lines are the "backbone" of the power grid system. They span long distances and traverse various terrains. They have the characteristics of wide coverage, harsh deployment environment, and high difficulty in operation and maintenance. Traditional manual inspection methods and installation of cameras and sensors can no longer meet the needs of the rapid development of the power grid. There is an urgent need to use new technologies to carry out refined inspections of transmission lines.

[0004] Currently, drone technology is widely used for detailed inspections. Since the large-scale application of drone technology, transmission line drone inspections have primarily been conducted through manual and autonomous methods. Manual inspections primarily involve pilots manually inspecting target towers and are applicable to all voltage levels. Autonomous inspections involve loading a route planned using laser point cloud technology and automatically flying to the target tower for point-by-point inspections. This is applicable to transmission lines of 220kV and above. However, laser point cloud technology suffers from high cost, low efficiency, a high barrier to entry, and poor flexibility. Consequently, grassroots teams conducting transmission line inspections face challenges such as high operational requirements, cumbersome defect identification, and difficulty inspecting channels. Consequently, existing technologies have begun to incorporate drones for autonomous inspections. However, the images captured by drones during autonomous inspections suffer from issues such as inaccurate flight altitudes, inability to accurately avoid sparse obstacles, misaligned images of the inspection target, inappropriate image brightness, and significant environmental influences. Summary of the Invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the present invention proposes a method and system for adaptive inspection of transmission lines. Based on lightweight artificial intelligence technology, it makes full use of existing hardware equipment and is deployed in the form of an operation terminal APP. It integrates target recognition, visual tracking, automatic obstacle avoidance, path correction, adaptive dimming, automatic focusing, multi-angle shooting and other algorithms, and intelligently completes adaptive inspection of transmission line towers of 110kV and below voltage levels, automatically generates autonomous inspection routes, and provides the capabilities of "reducing difficulty, improving efficiency, and accurate identification", reducing operating requirements, increasing defect recognition speed, and improving inspection methods.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0007] In a first aspect, a method for adaptive inspection of a transmission line is disclosed, comprising:

[0008] Get real-time images from drones;

[0009] A lightweight visual navigation model is used to identify the tower based on the real-time image, and the UAV performs a terrestrial flight to the top of the tower based on the tower shape characteristics of the identified tower; wherein the terrestrial flight uses GPS and visual SLAM to independently solve the posture, and outputs the optimal estimated fusion height through extended Kalman filtering fusion;

[0010] Using an identification algorithm to obtain tower parameters for the tower;

[0011] An inspection waypoint plan is determined according to the tower parameter matching database, and inspection images of key tower equipment are taken according to the inspection waypoint plan. The inspection images are taken using intelligent auxiliary photography to adjust image exposure.

[0012] In a second aspect, a transmission line adaptive inspection system is disclosed, comprising:

[0013] The data acquisition module is configured to: acquire real-time images of the UAV;

[0014] A cruise flight module is configured to: identify a tower using a lightweight visual navigation model based on the real-time image, and the UAV performs a terrestrial flight to the top of the tower based on the tower shape characteristics of the identified tower; wherein the terrestrial flight uses GPS and visual SLAM to independently solve the pose, and outputs an optimal estimated fused height through extended Kalman filtering;

[0015] A parameter acquisition module is configured to: acquire tower parameters by using an identification algorithm on the tower;

[0016] The image shooting module is configured to: determine the inspection waypoint plan according to the tower parameter matching database, and shoot inspection images of key tower equipment according to the inspection waypoint plan, and use intelligent auxiliary photography to adjust the image exposure.

[0017] In a third aspect, an electronic device is disclosed, including a memory and a processor, and computer instructions stored in the memory and running on the processor. When the computer instructions are run by the processor, the steps of the above-mentioned transmission line adaptive inspection method are completed.

[0018] In a fourth aspect, a computer-readable storage medium is disclosed for storing computer instructions, which, when executed by a processor, complete the steps of the above-mentioned transmission line adaptive inspection method.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The application provides a power transmission line adaptive inspection method, which is based on light-weight artificial intelligence technology, fully utilizes existing hardware devices, and integrates target identification, visual tracking, automatic obstacle avoidance, path correction, adaptive light adjustment, automatic focusing, multi-angle shooting and other algorithms in the form of a job terminal APP deployment to intelligently complete adaptive inspection of 110kV and below voltage level power transmission line towers, automatically generate autonomous inspection routes, and provide the ability of "reducing difficulty, improving efficiency, and accurate identification".

[0021] The application combines GPS positioning and visual SLAM (simultaneous localization and mapping) technology to achieve centimeter-level height control and accurately control the flight height of the unmanned aerial vehicle. The auxiliary license plate with exposure adjustment solves the problems of overexposure and low image resolution in shooting pictures, and greatly improves the quality of visible light images, laying a foundation for accurate defect identification in the later stage.

[0022] Based on artificial intelligence self-learning algorithm, the power transmission adaptive technology can adapt to complex terrains such as plains, rivers, mountains and hills, and has successfully adapted to a total of 8 categories and 24 sub-categories of towers such as square towers, dry towers, door towers, cat head towers and drum towers, basically covering the mainstream tower shapes of 35kV and 110kV voltage levels. The APP is downloaded and installed in APK form, adapts to the DJI 3 series of power grid industry, and is compatible with mainstream unmanned aerial vehicle control platforms. The route planning process is intelligent and automatic, and a single base tower route can be planned within 5 minutes without additional processing.

[0023] Through artificial intelligence and adaptive inspection technology, the problems of high cost, low efficiency, high threshold and poor flexibility of laser point cloud technology are completely overcome, helping power transmission and inspection front-line personnel to quickly solve the problems of autonomous route planning difficulty, high operation threshold, slow coordinate calibration progress, poor photo quality, and low defect identification accuracy, so that future 110kV and below voltage level power transmission line inspection has the obvious advantages of low difficulty, high efficiency and accurate identification.

[0024] The advantages of the additional aspects of the application will be partially given in the following description, partially will become obvious from the following description, or will be known by the practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0025] The drawings accompanying the specification of the application form part of the application and serve to provide further understanding of the application. The illustrative embodiments of the application and their description serve to explain the application without forming an improper limitation of the application.

[0026] Figure 1 The power transmission line adaptive inspection method flowchart described in embodiment one of the application.

[0027] Figure 2 The tower parameter acquisition schematic diagram described in embodiment one of the application.

[0028] Figure 3 A preliminary planning point diagram as described in embodiment one of the present application. DETAILED DESCRIPTION

[0029] It should be noted that the following detailed description is exemplary in nature and is intended to provide further description of the application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0030] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments in accordance with the present application.

[0031] In the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0032] Embodiment one

[0033] In one or more embodiments, a power line adaptive inspection method is disclosed, as shown, comprising the following steps: Figure 1

[0034] Step S1, acquiring real-time images of a UAV.

[0035] Step S2, determining and identifying a tower according to the real-time images using a lightweight visual navigation model, and the UAV flies to the top of the tower according to the tower shape features of the identified tower using a ground simulation flight.

[0036] Based on the lightweight front-end algorithm fusion visual navigation technology, the visual navigation technology is used to guide the UAV to quickly match the eight categories of tower types that have been adapted, and according to the corresponding tower shape features, the UAV is intelligently guided to fly to above the tower.

[0037] It can be understood that the eight categories of tower types include the upper character type, the rich character type, the drum type, the goat horn type, the cat head type, the wine cup type, the dry character type, and the door dry character type.

[0038] Through the feature key recognition technology, the redundancy capability of the algorithm model in a complex environment is improved, the sensitivity of the recognition algorithm to environmental changes is reduced, and the UAV is guided to autonomously fit the identified tower shape and intelligently move.

[0039] Step S2-1, automatically finding a tower according to the real-time images;

[0040] ​First, target tower identification: when adaptive inspection, the lightweight visual navigation model YOLO-FastestV2 is used to detect the tower feature, automatically identify the tower body, insulator string and conductor, associate the semantic label to the SLAM map node, and according to the identified target equipment and the direction of the transmission conductor, the position of the next base tower is obtained, and the calculation amount of pure geometric matching is reduced.

[0041] Preferably, in this embodiment, a multi-branch line can be selected: when adaptive inspection, if there are multiple branch routes in the SLAM map for the identified target tower, a list of all routes can be provided for selection, and at this time the pilot needs to manually slide (or click) to select the branch route to be flown.

[0042] Preferably, after determining the flight route to the target tower, adaptive obstacle avoidance is implemented during flight: when adaptive inspection, after detecting large-area obstacles such as trees, the flight strategy is dynamically adjusted based on adaptive obstacle avoidance technology to avoid tree obstacles, further ensuring the safety of unmanned aerial vehicle flight operation.

[0043] Specifically, based on the UAV image transmission picture, a target detection model is used for real-time image analysis to obtain a target detection result, the convex body coordinates of the obstacle are obtained according to the field of view angle of the UAV flight, the convex polyhedron coordinates covering the entire volume of the obstacle are obtained, and the corresponding coordinate information is taken as the edge of the obstacle; the distance and conversion angle between the UAV and the obstacle are obtained according to the position information of the UAV and the position information of the obstacle; the track points on the UAV flight route are screened according to the distance and conversion angle, and a set of track points capable of flying over the obstacle is obtained; the best track point is obtained by combining the smoothness summarized during the UAV flight process, taking the maximum turning angle of the UAV as the screening condition, and combining the expression for obtaining the optimal track point; and the optimal detour path is automatically calculated.

[0044] Through the above obstacle detection method, the embodiment can identify the entire occupied area of the hollow obstacle (such as sparse branches), help the UAV to bypass from the outside of the entire obstacle, avoid the problem that the UAV cannot pass through a small space, and improve the adaptability to complex environments.

[0045] The front-end AI vision technology is used to detect the obstacles (such as trees, buildings, etc.) around the transmission line in real time, the obstacle contour and distance are identified through a deep learning algorithm, the flight path and height of the UAV are dynamically adjusted to avoid collision risks. Compared with the traditional laser radar scheme, there is no need to scan the environment first, which reduces the operation preparation time. It can deal with sudden obstacles (such as temporary construction equipment) and improve flight safety.

[0046] The target identification and tracking algorithm is used to realize target tower identification, multi-branch line positioning and adaptive obstacle avoidance, so that the UAV flies along the transmission line, and the safety, efficiency and stability of the UAV inspection work are ensured.

[0047] Step S2-2, tower equipment automatic identification;

[0048] Through the equipment identification algorithm, the cross arm, insulator, hardware, clamp and other tower equipment are identified, the tower top, tower head, tower body, tower foundation and other tower body are identified, the ground wire, conductor and other line equipment are identified, and the key target equipment is focused on by the inspection personnel.

[0049] In this embodiment, the equipment identification algorithm uses a trained lightweight model to detect target equipment, and the equipment identification algorithm includes a four-level backbone network unit for extracting features, a spatial pyramid pooling unit (SPP) for extracting fixed size features, and a detection head for outputting results.

[0050] Step S2-3, flight simulation: through visual navigation, the height of the tower and the terrain undulation are calculated in real time, and the flight height of the unmanned aerial vehicle is automatically adjusted along the line to ensure stable inspection in complex terrain such as hills and mountains.

[0051] Specifically, combined with GPS positioning and visual SLAM (simultaneous localization and mapping) technology, centimeter-level height control is realized;

[0052] GPS and visual SLAM independently solve the pose, and the optimal estimate is output through extended Kalman filtering (EKF) fusion, including:

[0053] The GPS positioning obtains a first height, and the visual SLAM positioning obtains a second height, and the first height and the second height are preprocessed to obtain values in the same coordinate system;

[0054] The state variable is defined, which usually includes the height and its change rate:

[0055]

[0056] Wherein, x k is the state variable, h k is the fusion height of the unmanned aerial vehicle at time k, is the height change rate;

[0057] The system model, i.e. the state transition equation, is established:

[0058]

[0059] Where, Δt is the sampling time interval; w k is the process noise, which is subject to Gaussian distribution w k ~N(0,Q), Q is the process noise covariance matrix;

[0060] The observation model is established, and the first height and the second height are weighted and fused:

[0061]

[0062] Among them, v k is the observation noise, which obeys the Gaussian distribution v k ~N(0,R), R is the observation noise covariance matrix, and R GPS is the GPS observation noise, R SLAM is the SLAM observation noise.

[0063] Since both the system model and the observation model are linear, the Jacobian matrix of the EKF is simplified as follows:

[0064] State transition matrix:

[0065]

[0066] Observation matrix:

[0067]

[0068] Forecast states and covariances:

[0069]

[0070] in, is the optimal estimate of the state x at time k based on the observation data under the k-1 time condition;

[0071]

[0072] Among them, P k|k-1 is the covariance estimate, P k-1|k-1 is the optimal covariance estimate combined with the observed data at time k-1, and Q is the process noise;

[0073] Update EKF:

[0074]

[0075] Among them, K k is the Kalman gain, R is the observation noise, and The R matrix is ​​set according to the performance of GPS and SLAM. If the GPS noise is large, it will rely more on SLAM.

[0076]

[0077] in, is the state update at time k;

[0078] P k|k =(IK k Hk )P k|k-1

[0079] Among them, P k|k is the covariance update at time k;

[0080] The final fusion height is:

[0081]

[0082] in, is the first item of the updated state vector, h fusion is the final fusion height.

[0083] This embodiment uses terrain-simulating flight technology to avoid the risk of collision with the line caused by traditional fixed-altitude flight, reduce the frequency of manual adjustments by the pilot, and lower the probability of operational errors; at the same time, the combination of GPS and visual SLAM technology makes the flight altitude more accurate.

[0084] Step S3: using an identification algorithm to obtain tower parameters.

[0085] like Figure 2 As shown, the target recognition algorithm is used to complete the collection of longitude and latitude coordinate data of the center point of the tower, and the image recognition algorithm is used to complete quantitative data such as tower height, tower altitude, and tower head height, to assist inspection personnel in completing the basic ledger data management and update.

[0086] Specifically, according to the tower shape and the real-time image of the drone obtained in step S2, the longitude and latitude of the tower, the tower orientation angle, the line angle, the cross arm length, the tower height, and the tower head height are determined;

[0087] Among them, the tower latitude and longitude, and the tower orientation angle can be identified first and then manually confirmed;

[0088] Specifically, the longitude and latitude of the tower and the tower orientation angle are obtained from the GPS information of the tower, and the line angle, cross arm length, tower height and tower head height are matched with the corresponding data by the identified tower type results.

[0089] Step S4: determining an inspection waypoint plan based on the tower parameter matching database, and taking inspection images of key tower equipment according to the inspection waypoint plan, wherein the inspection images are taken using intelligent auxiliary photography to adjust image exposure.

[0090] Step S4-1: Preliminary waypoint planning;

[0091] Referencing the requirements of the State Grid Transmission Inspection Manual, combined with image recognition algorithms, it gradually identifies key equipment on towers, automatically selects the best shooting point and angle, intelligently plans routes and solves format conversion problems, and perfectly supports management requirements such as uploading, downloading, and reproducing planned routes.

[0092] Based on the tower parameters (fusible PMS3.0 account data) obtained in step S3, a preliminary flight route for autonomous flight is planned. As shown in the figure, the preliminary planning point is to guide the UAV to a general position, where the target component can be automatically photographed by AI adaptation. Planning the preliminary point requires the layer distance parameter of each layer cross arm, which has a certain error through AI recognition, so the UAV can be fine-tuned through AI control for data collection, which may take a long time. In order to improve efficiency, we should fully reuse the data of the previous tower when collecting data from the same line of sub-tower, especially for the same type of tower. The other side of the double-circuit symmetrical tower should also reuse the data. Figure 3

[0093] The UAV flies according to the preset flight route, and the flight instructions of the UAV and the angle of the information collection device are determined according to the tower parameters obtained in step S3 and the patrol flight point planning database;

[0094] The flight route generation technology can quickly determine the patrol flight route rules according to the adapted tower shape characteristics, and automatically plan the tower photographing points and flight safety points in combination with the set parameters to intelligently generate the fine patrol flight route.

[0095] Step S4-2: Adaptive adjustment of photographing target;

[0096] Based on the preliminary planning flight point, AI is used for automatic photographing of the three-section insulator, and when the position of the target to be detected is adjusted to the image data center area corresponding to the information collection device, the target component is photographed. The insulator of the tension tower will have horizontal and vertical angles due to the line corner and the height difference of the line tower, so special attention is needed.

[0097] Preferably, the AI adaptive photographing process can be based on the zooming method.

[0098] Preferably, the embodiment proposes visible light AI assisted photographing, and the specific process is as follows:

[0099] Step 1. Aperture index adjustment:

[0100] In this embodiment, the DJI Zen series camera used is a non-constant aperture lens, that is, as the zoom ratio changes, the maximum aperture value that can be set will also change, and the zoom ratios 2, 3, 4, and 5 correspond to 2.8, 3.2, 3.2, and 3.5, respectively.

[0101] Preferably, the maximum aperture at 2, 3, and 4 times zoom is set to 3.2.

[0102] Step 2. Parameter initialization, shutter speed is set to 500, and ISO is set to 100;

[0103] ​After locking the aperture, adjust the shutter and ISO, follow the principles:

[0104] (1) The smaller the shutter time (the larger the value) is better, and the maximum does not exceed the safe shutter speed;

[0105] (2) The smaller the ISO is better, and the noise is within the preset range.

[0106] Step 3. Adjust the brightness parameter according to the exposure. If the picture is too bright, dim it; if the picture is too dark, brighten it; until the exposure is normal, end the AI assisted shooting.

[0107] Wherein, the picture is too bright or too dark is set by the staff according to experience.

[0108] Through the AI assisted image enhancement algorithm, the problems of target object deviation, overexposure, low image resolution and other problems in the process of visible light image shooting are solved, which can greatly improve the quality of visible light image and lay a foundation for accurate identification of defects in the later period.

[0109] It should be noted that the three parameters that affect the exposure of the picture are aperture, shutter and ISO. Among them, the aperture affects the exposure and depth of field; the shutter affects the exposure and dynamic blur; ISO affects the exposure and noise. In this embodiment, the focus is affected by the depth of field, so the adaptive focusing process needs to determine the aperture value first.

[0110] Step S4-3: Record the shooting point in the adaptive shooting process, and output the complete route at the end of the process.

[0111] The adaptive planning route proposed in this embodiment only needs the operation and maintenance personnel to carry out the work of the airplane one key takeoff, through the visual guidance algorithm, control the unmanned aerial vehicle to fly to the tower position; through fast and accurate identification and correction, generate the patrol route this time after executing shooting. Except for the cost of APP / algorithm updating and optimization, no additional investment is needed. Relying on various parameters collected in the adaptive patrol process, the tower shooting point and the flight safety point are automatically planned, and the power transmission fine patrol route is intelligently generated.

[0112] Embodiment two

[0113] In one or more embodiments, a power transmission line adaptive inspection system is disclosed, specifically comprising:

[0114] The data acquisition module is configured to acquire real-time images of the unmanned aerial vehicle;

[0115] The cruise flight module is configured to determine an identification tower according to the real-time image by using a lightweight visual navigation model, and the unmanned aerial vehicle flies to the top of the tower according to the tower shape feature of the identified tower by using the ground effect flight.

[0116] The parameter acquisition module is configured to acquire tower parameters of the tower by using an identification algorithm.

[0117] The image shooting module is configured to determine an inspection flight point plan according to the tower parameters matched with a database, and shoot inspection images of key equipment of the tower according to the inspection flight point plan.

[0118] Embodiment three

[0119] The embodiment provides an electronic device, which comprises a memory and a processor, and computer instructions stored in the memory and running on the processor, and when the computer instructions are run by the processor, the steps of the power transmission line adaptive inspection method are completed.

[0120] Embodiment four

[0121] The embodiment provides a computer readable storage medium for storing computer instructions, and when the computer instructions are executed by a processor, the steps of the power transmission line adaptive inspection method are completed.

[0122] The present application is described with reference to flowcharts and / or block diagrams of the method, device (system) and computer program product according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device that realizes the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0123] These computer program instructions can also be stored in a computer readable memory capable of guiding the computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer readable memory produce a product including instruction devices, which realize the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device, and a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide the functions for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0125] The descriptions of the various embodiments in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0126] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A transmission line adaptive inspection method, characterized in that: include: Get real-time images from drones; A lightweight visual navigation model is used to identify the tower based on the real-time image, and the UAV performs a terrestrial flight to the top of the tower based on the tower shape characteristics of the identified tower; wherein the terrestrial flight uses GPS and visual SLAM to independently solve the posture, and outputs the optimal estimated fusion height through extended Kalman filtering fusion; Using an identification algorithm to obtain tower parameters for the tower; An inspection waypoint plan is determined according to the tower parameter matching database, and inspection images of key tower equipment are taken according to the inspection waypoint plan. The inspection images are taken using intelligent auxiliary photography to adjust image exposure.

2. A transmission line adaptive inspection method according to claim 1, characterized in that: The lightweight visual navigation model YOLO-FastestV2 is used to detect tower features and identify towers. The semantic labels are associated with SLAM map nodes. The orientation of the next tower is determined based on the identified target equipment and the direction of the transmission line, and the inspection route is obtained.

3. The method for adaptive inspection of a power transmission line according to claim 1, wherein: Through visual navigation, the tower height and terrain undulations are calculated in real time, and the drone is controlled to automatically adjust its flight altitude along the route. Specifically, GPS positioning and visual SLAM technology are combined to achieve altitude control of the UAV.

4. A transmission line adaptive inspection method according to claim 3, characterized in that: The GPS and visual SLAM independently solve the pose and output the optimal estimated fusion height through extended Kalman filter fusion, including: The first height is obtained by GPS positioning, and the second height is obtained by visual SLAM positioning. The first and second heights are preprocessed to obtain values ​​in the same coordinate system; Weighted fusion of the first height and the second height: Among them, v k is the observation noise, which obeys the Gaussian distribution v k ~N(0,R), where R is the observation noise covariance matrix.

5. The method for adaptive inspection of a power transmission line according to claim 1, wherein: The inspection waypoint planning is determined according to the tower parameters. Specifically, the UAV matches the preset route and flight instructions according to the identified tower parameters.

6. The method for adaptive inspection of a power transmission line according to claim 1, wherein: When the position of the target to be detected is adjusted to the center of the image, the target component is photographed; The shooting points are recorded during the adaptive shooting process, and the complete route is output when the process is completed.

7. A transmission line adaptive inspection method according to claim 6, characterized in that: The intelligent assisted photography adjusts the image exposure as follows: Aperture index adjustment and confirmation; Initialize the shutter and ISO parameters, and optimize them according to the rules of reducing shutter time and ISO; Adjust the brightness parameters according to the exposure situation; if the picture is too bright, dim it; if the picture is too dark, brighten it; Until the exposure is normal, end the AI-assisted photography.

8. A transmission line adaptive inspection system, characterized in that: include: The data acquisition module is configured to: acquire real-time images of the UAV; A cruise flight module is configured to: identify a tower using a lightweight visual navigation model based on the real-time image, and the UAV performs a terrestrial flight to the top of the tower based on the tower shape characteristics of the identified tower; wherein the terrestrial flight uses GPS and visual SLAM to independently solve the pose, and outputs an optimal estimated fused height through extended Kalman filtering; A parameter acquisition module is configured to: acquire tower parameters by using an identification algorithm on the tower; The image shooting module is configured to: determine the inspection waypoint plan according to the tower parameter matching database, and shoot inspection images of key tower equipment according to the inspection waypoint plan, and use intelligent auxiliary photography to adjust the image exposure.

9. An electronic device, characterized in that: The method comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the method for adaptive inspection of power transmission lines according to any one of claims 1 to 7 is completed.

10. A computer-readable storage medium, characterized in that Used to store computer instructions, which, when executed by a processor, complete the transmission line adaptive inspection method according to any one of claims 1 to 7.

Images