System and method for drone patrol inspection of meshed dock

The meshed dock system with optimized route planning, double-restraint technology, AI integration, and precision landing addresses inefficiencies in drone patrol inspections, enhancing coverage, stability, and safety while reducing costs.

JP2024542562AActive Publication Date: 2024-11-15STATE GRID INTELLIGENCE TECHNOLOGY CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
JP2024531210
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-12-03
Filing Date
2022-11-02
Publication Date
2024-11-15
Estimated Expiration
2042-11-02

AI Technical Summary

Technical Problem

Existing drone patrol inspection systems face issues with incomplete coverage, complex and prone-to-failure battery replacement mechanisms, low intelligence leading to subjective judgment errors, inaccurate landing, and inefficient route planning, resulting in increased costs and reduced safety.

Method used

A meshed dock system with optimized route planning, double-restraint technology for stable drone landing, integrated AI for accurate judgment, and precision landing using RTK positioning and visual tracking, along with a method for efficient battery replacement.

Benefits of technology

Enhances coverage efficiency, reduces labor costs, improves stability and safety, and ensures accurate and efficient drone patrol inspections by minimizing inspection time and optimizing routes.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024542562000001_ABST
    Figure 2024542562000001_ABST
Patent Text Reader

Abstract

A system and method for patrol inspection of a drone (33) on a meshed dock (3, 34) including a plurality of docks (3, 34) arranged in a mesh and each used to accommodate at least one drone (33), wherein the dock (3, 34) includes a dock controller communicating with a control terminal, the dock controller communicating with a drone remote control, the drone remote control communicating with the drone (33), and the control terminal is used to generate an optimal patrol inspection route for the drone (33) based on the current flight range of the drone (33) and the distance from the patrol inspection target to each dock (3, 34) with the optimization goal of minimizing the patrol inspection time, and send it to the dock controller, with beneficial effects being to realize efficient cooperative patrol inspection of the drone (33) based on the drone docks (3, 34) arranged in a mesh, reduce labor costs, and meet the needs of regular or emergency patrol inspection of a plurality of patrol inspection targets across a plurality of different fields.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to the technical field of power patrol inspection, and in particular to a system and method for drone patrol inspection of a meshed dock. [Background technology]

[0002] The description in this section merely provides background art related to the present invention and does not necessarily constitute prior art.

[0003] The drone dock serves as a security relay station for drones, and its location is extremely important for the drones, directly affecting the radius, work efficiency and results of the drone's flight patrol inspections.

[0004] The inventors have discovered that the existing power patrol inspection methods have the following problems.

[0005] (1) The location of the drone dock is set freely, or only a few drone docks are sparsely placed, which often makes it impossible to fully cover the targets waiting for patrol inspection. Furthermore, during the patrol inspection process, a single flight patrol inspection is often performed for each target waiting for patrol inspection in sequence, which does not cover the nearby waypoints, resulting in a situation where the flight path and power during the patrol inspection are wasted.

[0006] (2) Existing drone docks use a large number of robotic arms to control drone centering and battery replacement, making battery replacement cumbersome. Furthermore, the coordination between the drone and the multi-degree-of-freedom robotic arm or battery replacement mechanism is prone to causing malfunctions in the robotic arm or drone, which in turn causes damage to the equipment and reduces the stability of the entire system.

[0007] (3) The drone intelligent dock can achieve mutual communication with the background monitoring center. After monitoring the flight environment on-site, the operators at the background monitoring center often make subjective judgments of the on-site flight conditions, which are low in intelligence, highly subjective, and have a certain possibility of misjudgment; or the judgments on the flight conditions are simply collected and lack of comprehensive judgment of multi-source data, which brings certain safety concerns to flight tasks.

[0008] (4) When a drone identifies the coordinates of a waiting position for landing, highly accurate real-time positioning is required. Not only is the cost of positioning components high, but it also requires obtaining coordinate data of the position of a preset point on the drone's airport in real time, making landing control complicated. In the prior art, there is a method of achieving landing by identifying a specific image of the landing point, but in most cases only a single data source is identified, and landing accuracy cannot be guaranteed.

[0009] (5) In the prior art, a method has been disclosed for utilizing binocular vision to hover and perform positioning and ranging. However, the drone is still controlled to reach inspection targets in sequence, brake and hover, take a photograph at a designated point, and then accelerate to the next inspection target. The braking, hovering, and acceleration of the drone consume a large amount of battery power, and it is not possible to perform autonomous inspection without hovering. Summary of the Invention [Problem to be solved by the invention]

[0010] In order to solve the shortcomings of the prior art, the present invention provides a system and method for drone patrol inspection of a meshed dock, which realizes efficient drone collaborative patrol inspection based on a meshed drone dock, reduces labor costs, and meets the regular or emergency inspection needs of multiple patrol inspection targets across multiple different fields. [Means for solving the problem]

[0011] In order to achieve the above object, the present invention employs the following technical aspects.

[0012] The first aspect of the present invention is a plurality of docks arranged in a meshed configuration, each dock being used to house at least one drone; The dock includes a dock controller in communication with the control terminal, the dock controller in communication with a drone remote control, the drone remote control in communication with the drone; The control terminal is used to optimize the inspection time by obtaining the inspection target corresponding to each dock based on the current range of the drone and the distance from the inspection target to each dock, and generate an optimal inspection route for each drone based on the determined inspection target and send it to the dock controller; We provide a drone patrol inspection system for meshed docks.

[0013] Furthermore, the dock The present invention relates to a dock body, and a mounting mechanism, a vertical fixing mechanism and a horizontal fixing mechanism provided in the dock body, the mounting mechanism including an extendable landing platform and a first motor, the landing platform being driven by the first motor; The vertical fixing mechanism includes a first centering rod having one end provided on a side wall of the dock body via a pivot shaft and a gear, and the landing platform is provided with a rack that engages with the gear, and the first centering rod is driven to rotate around the pivot shaft by the engagement between the gear and the rack; The lateral fixing mechanism includes a rotating rod whose both ends are provided on the side walls of the dock body, a second centering rod provided on the rotating rod, and a second motor, and the rotating rod is driven by the second motor to rotate along a direction opposite to the movement direction of the landing platform relative to the dock body and drive the second centering rod to move along a direction perpendicular to the movement direction of the landing platform.

[0014] Furthermore, as the gear and rack mesh, the first centering rods provided on both side walls rotate around their axes, causing the other ends of the two first centering rods to move to intermediate positions or to spread apart in both directions.

[0015] In addition, a second centering rod is provided at both ends of the rotating rod, and when the landing platform is driven and reset, the rotating rod rotates in the positive direction, and the second centering rods at both ends move along the rotating rod to an intermediate position to laterally restrain the drone; The meshing of the rack and gears moves the first centering rods on both side walls to an intermediate position, restraining the drone vertically.

[0016] In addition, a second centering rod is provided at both ends of the pivot rod, and when the landing platform is pushed out from the dock body, the pivot rod rotates in the opposite direction, and the second centering rods at both ends move to both sides along the pivot rod to release the lateral constraint on the drone; The engagement of the rack and the gear causes the first centering rods on both side walls to move apart to either side, releasing the vertical restraint on the drone.

[0017] Furthermore, the dock The dock includes a drone parking space, a charging module, and an energy storage module. The dock body is provided with an attachment module that secures the dock body with a screw-type automatic locking structure, and the drone parking space is provided with a drone fixing device that autonomously suppresses vibration in the horizontal and vertical directions, and the dock controller communicates with the charging module and the attachment module respectively.

[0018] In addition, the drone is equipped with a three-axis head, an RTK positioning module and a front-end AI processing module, A camera and a video camera are mounted on the three-axis pan head, the camera is a monocular camera capable of zooming, and the video camera is used to obtain video information of the tower / pillar, where the camera and the video camera are integrated into one lens.

[0019] The RTK positioning module is used to measure the drone's three-dimensional coordinate information. The front-end AI processing module is Fitting the drone's flight control data, the RTK positioning module data, and the images captured by the zoom camera, it issues flight control commands to control the drone's flight, controls the camera head to adjust the camera angle and zoom, locks onto the patrol inspection target and takes photos, Using a visual zoom wide-angle camera, take photos during the flight process approaching the hovering point, calculate the coordinate values ​​(GPS values) where the photos were taken and the attitude of the camera head, and identify the patrol inspection target in the photo according to the camera imaging principle. According to the current GPS position and three-dimensional speed of the drone, as well as the roll angle, pitch angle and yaw angle of the head attitude, the Kalman filter algorithm is used to adjust the position of the drone head, and the zoom camera is used to lock on to the target inspection point of the tower or pillar by zooming. It is configured to improve the accuracy of information collection of patrol inspection targets and the quality of collected images by completing information collection for target inspection viewpoints of towers and columns by taking photos.

[0020] A second aspect of the present invention is Obtaining a distance from a patrol inspection target to each dock; selecting a dock that is closest to the patrol inspection target as the optimal dock; Sequentially determining each patrol inspection target to obtain a patrol inspection target corresponding to each dock; Among them, each dock performs the planning task of the patrol inspection task, and the task is: numbering the patrol inspection targets based on the distance from the patrol inspection targets within the dock range to the dock, the greater the distance, the greater the number; For each patrol inspection target within the dock range, determine whether the difference between the total flight time of the drone and the time required for the patrol inspection target to perform one patrol inspection alone is smaller than the minimum time required for the other patrol inspection targets within the dock range to perform one patrol inspection alone, and if YES, set this patrol inspection target as a single tower task; If NO, set this patrol target as the current patrol target, and determine whether the sum of the time from the dock to the current patrol target, the patrol time of the current patrol target, the patrol time from the current patrol target to the nearest secondary patrol target whose number is smaller than that of the current patrol target, the time from the current patrol target to the secondary patrol target, and the time from the secondary patrol target to the dock is greater than the total flight time of the drone; if YES, set the current patrol target as a single tower task; if NO, execute the route task of two towers and pillars to sequentially perform patrol inspections of the current patrol target and the secondary patrol target. A method for drone patrol inspection of a meshed dock is provided.

[0021] A third aspect of the present invention is Using the drone patrol inspection system of the meshed dock mentioned above, Acquiring in-dock environment information and out-dock environment information within a sensing range; Select a target dock based on the drone's position, determine corresponding flight influence factors based on the flight command, obtain corresponding flight environment data from the external environment information of the selected target dock, and control the drone to return if the flight environment data does not satisfy the flight conditions; Determining corresponding landing influence factors according to a return command, obtaining corresponding landing environment data and return environment data from the outside environment information and the inside environment information of the target dock according to the landing influence factors, controlling the landing method of the drone according to the landing environment data until the drone returns to the target dock, and adjusting the inside environment of the target dock according to the return environment data. Provides a method for determining the drone task execution environment.

[0022] Selecting a target dock based on the drone position specifically includes determining the detection range of the dock where the drone is located based on the drone position, and determining the dock within the detection range as the target dock; and if two or more docks are both within the detection range, determining the dock closest to the drone as the target dock.

[0023] A fourth aspect of the present invention is Using the drone patrol inspection system of the meshed dock mentioned above, Obtaining drone positioning data; Based on the acquired positioning data, determine whether the drone is located within a preset landing range, and if the drone is not located within the preset landing range, control the drone to move until the position request is satisfied; After it is determined that the drone is located within the preset landing range, when the drone is located at a position that is a first preset distance away from the landing point, obtain image data or video data below the drone, and if the precision landing range code cannot be identified based on the obtained image data or video data, control the drone to descend to a position that is a third preset distance away from the landing point and identify the precision landing range code again until the precision landing range code is identified; if a precision landing range code is identified based on the acquired image data or video data, controlling the drone to descend to a position a second preset distance away from the landing point and again acquire image data or video data below the drone, and if a precision landing position code is identified based on the again acquired image data or video data, controlling the drone to descend to a position a fourth preset distance away from the landing point and controlling the drone to land. Provides a method for controlling precision landing of drones.

[0024] A fifth aspect of the present invention is Using the drone patrol inspection system of the meshed dock mentioned above, S1, which adopts an image acquisition module on the camera platform to obtain real-time wide-angle images of the patrol inspection target before the drone enters the detection point at a constant speed according to the patrol inspection request; It is determined whether the patrol inspection target is located in the captured real-time image, and if it is YES, proceed to step S3, and if it is NO, control the movement of the camera platform to change the attitude until the patrol inspection target is found in the real-time image in S2; The processing module adopts a Kalman filter algorithm to fit the drone's shooting position and the head attitude based on the information of the patrol inspection target position, the drone's shooting position, and the head attitude in the real-time image, and determines the focal length mode of the image acquisition module S3; Control the drone to fly at a constant speed to the shooting position obtained by fitting. During the flight, the processing module adjusts the attitude of the pan head in real time inversely according to the three-dimensional direction in which the drone flies at a constant speed, and adjusts the focal length mode of the image acquisition module S4, so as to achieve that the patrol inspection target is locked on to the setting area of ​​the real-time image of the image acquisition module; The drone arrives at the shooting position, confirms that the patrol inspection target position is within the setting area of ​​the real-time image of the image acquisition module, locks on the inspection viewpoint and acquires images S5; The processing module processes the captured picture, and controls the drone to perform the next detection point task, and re-executes S1 until completing the image capture task of all detection points. A drone patrol inspection method based on visual movement tracking is provided. Effect of the Invention

[0025] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0026] 1. The present invention innovatively designs a drone patrol inspection system for meshed docks and puts forward a drone patrol inspection method for meshed docks. Based on the current range of the drone and the distance from the patrol inspection targets to each dock, the optimization goal is to minimize the patrol inspection time. Multiple patrol inspection targets corresponding to each dock are obtained. Based on the determined patrol inspection targets, an optimal patrol inspection route for each drone is generated, solving the optimization problem of collaborative patrol inspection of single-tower tasks and multi-tower tasks, realizing the optimization of patrol inspection routes with the goal of the shortest patrol inspection time, realizing efficient collaborative drone patrol inspection based on meshed drone docks for patrol inspection targets in various fields, reducing labor costs, and meeting the regular or emergency patrol inspection needs of multiple patrol inspection targets in various fields.

[0027] 2. The present invention innovatively proposes a drone dock, designs a double restraint technology of lateral and vertical restraint for the drone, and proposes a method of combining a centering rod set fixing mechanism and a gear rack mechanism to center the drone, solving the problems of the single-scene limitations of the drone dock and the stability of drone parking, and improving the landing stability of drones in different patrol inspection environments. Moreover, as a general-purpose dock, the drone dock supports remote control operations, significantly improves the efficiency of patrol inspection work, diversifies application scenarios, and enables drones to cover a larger range.

[0028] 3. The present invention innovatively puts forward a method for judging the task execution environment of a drone, by combining different flight tasks or return tasks with different task environmental conditions and respectively judging whether they are suitable for performing the tasks, thereby meeting the needs of the judgment logic of the dock flight conditions, and realizing redundancy of judgment conclusions through the dock self-judgment method in complex flight situations, improving the accuracy of judgment, solving the limitations of existing flight environment monitoring technology that the judgment conditions are single, subjective interference, and low intelligence, and no manual intervention is required, realizing autonomous prediction of flight conditions for different tasks, and significantly improving the efficiency of drone patrol inspection and the safety of the dock system.

[0029] 4. The present invention innovatively proposes a control method for precision landing of a drone, which realizes initial location of the drone and the waiting position for landing based on the drone's positioning data, and combines real-time differential positioning data, precision landing range code and precision landing position code, solves the problem of difficulty in controlling precision landing of a drone due to constant image recognition and distance approach, realizes precise and step-by-step control of drone landing, and improves the control accuracy of drone landing.

[0030] 5. The present invention innovatively puts forward a drone patrol inspection method based on visual movement tracking. During the flight process from when the drone enters the patrol inspection target to when it leaves the patrol inspection target, the drone always flies along a set flight path, and uses a Kalman filter algorithm to fit the current position and speed data to adjust the head attitude and camera zoom in real time, thereby realizing the camera's movement tracking and lock-on shooting of the patrol inspection target. The drone can automatically capture images of the patrol inspection target during the patrol inspection process without hovering, greatly reducing the labor intensity of the patrol inspection workers. Furthermore, the present invention adopts a reverse movement tracking method, and dynamically adjusts the attitude of the drone and the camera on the head to achieve relative stillness with the patrol inspection target, greatly saving the power consumption of the drone and the workload of one flight. The acquisition of the patrol inspection target of the present invention is completed based on a monocular camera, which has a simple structure and low cost.

[0031] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

[0032] The accompanying drawings of the specification which form a part of the present invention are intended to provide a further understanding of the present invention, and the schematic embodiments of the present invention and the description thereof are intended to interpret the present invention and are not intended to unduly limit the present invention. [Brief description of the drawings]

[0033] [Figure 1]FIG. 1 is a schematic diagram of a drone patrol inspection system for a meshed dock according to a first embodiment of the present invention. [Diagram 2] 1 is a schematic diagram of a drone dock according to Example 1 of the present invention. [Diagram 3] FIG. 1 is a schematic diagram of a drone dock body according to a first embodiment of the present invention. [Figure 4] FIG. 4 is a schematic diagram of the drive of the gear at the bottom of the first centering rod according to the first embodiment of the present invention. [Diagram 5] FIG. 4 is a schematic diagram showing the resetting of the first centering rod and the second centering rod according to the first embodiment of the present invention. [Figure 6(a)] FIG. 2 is a schematic diagram showing centering of the drone according to the first embodiment of the present invention. [Figure 6(b)] FIG. 2 is a schematic diagram showing centering of the drone according to the first embodiment of the present invention. [Figure 7(a)] FIG. 2 is a schematic diagram showing the landing of a drone according to the first embodiment of the present invention. [Figure 7(b)] FIG. 2 is a schematic diagram showing the landing of a drone according to the first embodiment of the present invention. [Figure 8] FIG. 1 is a schematic diagram illustrating an installation of a drone dock according to a first embodiment of the present invention. [Figure 9(a)] FIG. 11 is a schematic diagram of the internal structure of a mobile drone dock according to a second embodiment of the present invention. [Figure 9(b)] FIG. 11 is a structural schematic diagram of the door of the mobile drone dock according to the second embodiment of the present invention. [Figure 10] FIG. 6 is a structural schematic diagram of a charging module according to Example 2 of the present invention. [Figure 11] FIG. 11 is a structural schematic diagram of a fixing device for a drone parking space according to Example 2 of the present invention. [Figure 12] FIG. 11 is a schematic diagram of the local structure of a fixing device for a drone parking space according to Example 2 of the present invention. [Figure 13(a)] FIG. 6 is a structural schematic diagram of a mounting module according to a second embodiment of the present invention. [Figure 13(b)] FIG. 11 is a schematic diagram of a local structure of an attachment module according to a second embodiment of the present invention. [Figure 13(c)]FIG. 11 is a schematic diagram of a local structure of an attachment module according to a second embodiment of the present invention. [Figure 14] 11 is a work flow chart of a drone system according to Example 2 of the present invention. [Figure 15] FIG. 11 is a schematic diagram of a route planning flow according to a third embodiment of the present invention. [Figure 16] FIG. 1 is a schematic diagram 1 of a task plan according to a third embodiment of the present invention. [Figure 17] FIG. 2 is a schematic diagram 2 of a task plan according to the third embodiment of the present invention. [Figure 18] FIG. 11 is a schematic diagram of task commands and corresponding environmental factors according to a fourth embodiment of the present invention; [Figure 19] FIG. 11 is a schematic diagram of a determination of the execution environment of a storage task of a drone according to Example 4 of the present invention. [Figure 20] 11 is a schematic diagram of a flow of a control method for precision landing of a drone according to Example 5 of the present invention. [Figure 21] FIG. 13 is a schematic diagram showing a flow of an autonomous patrol inspection method for a drone according to a sixth embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0034] The present invention will now be further described with reference to the drawings and examples.

[0035] It should be noted that the following detailed description is illustrative and is intended to provide further explanation for the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art.

[0036] It should be noted that the terms used herein are merely for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments of the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form, and it should be further understood that when the terms "comprise" and / or "comprise" are used herein, they denote the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0037] Unless there is a conflict, the embodiments of the present invention and the features of the embodiments can be combined with each other.

[0038] Example 1 As shown in FIG. 1 , the first embodiment of the present invention includes a plurality of docks 34 arranged in a meshed configuration, each of which is used to house at least one drone 33; The dock includes a dock controller in communication with the control terminal, the dock controller in communication with a drone remote control, the drone remote control in communication with the drone; The control terminal provides a meshed dock drone patrol inspection system, which is used to optimize the inspection time by obtaining patrol inspection targets corresponding to each dock based on the drone's current range and the distance from the patrol inspection targets to each dock, generate optimal patrol inspection routes for each drone based on the determined patrol inspection targets, and send them to the dock controller.

[0039] In this embodiment, the substation is selected as the main placement point of the drone dock, and the drone can first patrol and inspect the substation equipment nearby. The area around the substation is the main junction of power lines and is also an area that requires focused inspection. By patrolling around the substation, the drone can maximally inspect the power lines. The drone dock is placed inside the substation and can be used as part of the inspection and repair of the substation, making it easy to operate and maintain.

[0040] In some other embodiments, the drone dock may be appropriately deployed on land such as a 5G base station or a mountaintop solar power plant, and may be used as a patrol inspection target for targets in other fields, such as communications, firefighting, etc., and it can be understood that the drone dock can be deployed as long as there is electricity.

[0041] As shown in FIGS. 2 and 3, the drone dock described in this embodiment is a miniaturized drone dock including a dock body, a mounting mechanism, a vertical fixing mechanism and a horizontal fixing mechanism provided in the dock body, and the mounting mechanism includes an extendable landing platform and a first motor, and the landing platform is driven by the first motor; The vertical fixing mechanism includes a first centering rod having one end provided on a side wall of the dock body via a pivot shaft and a gear, and the landing platform is provided with a rack that engages with the gear, and the first centering rod is driven to rotate around the pivot shaft by the engagement between the gear and the rack; The lateral fixing mechanism includes a rotating rod whose both ends are provided on the side walls of the dock body, a second centering rod provided on the rotating rod, and a second motor, and the rotating rod is driven by the second motor to rotate along a direction opposite to the movement direction of the landing platform relative to the dock body and drive the second centering rod to move along a direction perpendicular to the movement direction of the landing platform.

[0042] In this embodiment, the dock body 6 has a rectangular frame structure, and a top cover 5 is provided at the top end of the dock body 6. A solar power generation panel is provided on the top cover 5. The solar power generation panel absorbs light energy, converts the light energy into electrical energy and stores it to provide power for the dock.

[0043] Preferably, the top cover 5 is of a sloping design to prevent water from pooling on the top of the dock.

[0044] In this embodiment, an extendable landing platform 4 is provided in the dock body 6. When the drone lands, the landing platform 4 is pushed out from inside the dock body 6 to place the drone on it. After the drone lands, the landing platform 4 is retrieved into the dock body 6. When the drone performs a patrol inspection task, the landing platform 4 is pushed out from inside the dock body 6, the drone takes off, and then the landing platform 4 is retrieved into the dock body 6.

[0045] In this embodiment, three sides of the dock body 6 are closed, and the forward side and the landing platform 4 form a closed surface, which ensures the protective performance of the whole dock.

[0046] In this embodiment, the first motor 9 is connected to the landing platform 4 via a rod, and drives the landing platform 4 to be pushed out of the dock body 6 or to be retrieved into the dock body 6 .

[0047] Preferably, two first motors 9 are provided.

[0048] In this embodiment, as shown in FIG. 3, slide rails are provided at both ends of the dock body 6, and both ends of the rotating rod 7 are mounted on the slide rails of the dock body 6 by rolling pulleys, and the rotation is controlled by a second motor 8, and the rotation direction of the rotating rod 7 is opposite to the moving direction of the landing platform 4.

[0049] A slide groove is provided at the end of the second centering rod 10, and one second centering rod 10 is provided at each end of the rotating rod 7. The second centering rod 10 is attached to the rotating rod 7 via the slide groove and moves in one direction along the direction of the rotating rod 7 as the rotating rod 7 rotates, i.e., moves laterally along a direction perpendicular to the rotation direction of the rotating rod 7.

[0050] The rotating rod 7 is provided with a screw thread, and the second centering rod 10 moves in one direction along the screw thread as the rotating rod 7 rotates via a slide groove.

[0051] Preferably, the pivot rod 7 adopts a screw.

[0052] The rotating rod 7 rotates on the slide rail by being driven by a second motor 8, and the direction of movement of a second centering rod 10 is controlled based on the rotating direction of the rotating rod 7. Preferably, when the rotating rod 7 rotates in the forward direction, the second centering rods 10 on both sides perform a centering motion, i.e., move to an intermediate position, and when the rotating rod 7 rotates in the reverse direction, the second centering rods 10 on both sides move in the opposite direction, i.e., spread out to both sides, so that the rotation of the rotating rod 7 on the slide rail and the movement of the second centering rod 10 through the slide groove cooperate to balance the acting force of the second centering rod 10 reciprocating with the rotating rod 7, and ensure that the second centering rod 10 is displaced with a degree of freedom in one direction.

[0053] Preferably, when the landing platform 4 is pushed out, when the pivot rod 7 rotates in the opposite direction, the second centering rods 10 on both sides move in the opposite direction, i.e., spread to both sides, in this case also used for the drone on the landing platform 4 to fly out after the second centering rods 10 spread; When the landing platform is retrieved and reset, when the pivot rod 7 rotates in the positive direction, the second centering rods 10 on both sides perform a centering movement, i.e., move to the middle position, which is also used to reset the drone laterally on the landing platform and laterally restrain and fix the drone.

[0054] In this embodiment, a first centering rod 13 is provided on each of the two opposing side walls of the dock body 6, and the first centering rods 13 on both sides rotate around their axes through the engagement of the gear and the rack, thereby moving the other ends of the first centering rods 13 to an intermediate position or spreading out in both directions.

[0055] In this embodiment, as shown in FIG. 4, a gear is provided on the first centering rod 13, and a rack 14 is connected to the landing platform 4 via a screw, and the rack 14 meshes with the gear. When the landing platform is pushed out and returned, the gear rotates, and the meshing between the gear and the rack causes the first centering rod 13 to rotate around the axis. When the first centering rod 13 rotates around the axis, the moving force is converted into a rotational force moment through the transmission between the gear and the rack.

[0056] Preferably, the first centering rod 13 is used for vertical resetting of the drone, and the first centering rod 13 rotates circumferentially around the axis via a pivot shaft, and the other end rotates around the axis to an intermediate position to fix the drone.

[0057] In this embodiment, the reset of the drone is completed by pushing the second centering rod 10 and the first centering rod 13 together. As shown in FIG. 5, the reset of the drone is divided into two parts, one part is reset horizontally by pushing the second centering rod 10, and the other part is reset vertically by rotating the first centering rod 13.

[0058] Preferably, before the drone takes off, the first motor 9 pushes the landing platform 4 to open the front side of the dock body 6. During the opening process, the second centering rod 10 is expanded. Due to the rotation of the rotating rod 7 in the opposite direction, the second centering rods 10 on both sides move in the opposite direction, that is, expand to both sides, and release the lateral fixation on the drone. At the same time, the rack connected to the landing platform 4 is pushed forward in synchronization with the landing platform 4. Due to the meshing between the rack 14 on the landing platform 4 and the gear on the first centering rod 13, the first centering rod 13 on both sides is moved to rotate, and the first centering rods 13 on both sides also expand to both sides by rotating around the axis, and release the vertical fixation on the drone. Thus, the drone takes off autonomously according to the planned route and performs patrol inspection work.

[0059] After completing the patrol inspection task, the drone precisely lands on the landing platform 4 with visual aid, and then the first motor 9 moves the landing platform 4 to perform the cabin closing operation. In the process of closing the cabin door, the positive rotation of the rotating rod 7 causes the second centering rods 10 on both sides to perform a centering movement, i.e., to move to the middle position, completing the lateral reset of the drone; at the same time, the meshing between the rack 14 and the gear causes the first centering rod 13 to rotate, and the first centering rods 13 on both sides rotate around the axis, moving to the middle to fix the drone, completing the longitudinal reset of the drone. As shown in Figures 6(a)-6(b) and 7(a)-7(b), a schematic diagram of the centering and landing of the drone is shown.

[0060] In this embodiment, charging rods 11 are further provided on both ends of the dock body 6, and several charging ports 12 are provided on the charging rods. After the drone is reset, the charging contact plate at the bottom of the drone contacts the charging ports 12, and charging is performed according to the control command of the dock.

[0061] When the drone performs a patrol inspection task, it first detects the remaining power of the battery, and when the power is insufficient, it charges the drone's power battery through the charging port 12; when the drone's power is sufficient, it pushes out the landing platform 4 to allow the drone to take off.

[0062] In more embodiments, the above drone dock can be a universal drone dock that can be applied to towers and columns, and the installation process is as shown in Figure 8, in which the dock 3 is attached to the tower / column 1 by the dock bottom support 2, the dock 3 is connected to the dock bottom support by screws, and the dock bottom support 2 is fixed to the tower / column 1 by bolts. In different terrains, they can all be installed by relying on towers and columns, which can realize diversification of scenes.

[0063] In more embodiments, the drone dock can be used in conjunction with a vehicle-mounted drone, with the drone dock mounted to the roof by a dock bottom support.

[0064] Example 2 A second embodiment of the present invention includes a plurality of docks arranged in a meshed configuration, each dock being used to house at least one drone; The dock includes a dock controller in communication with the control terminal, the dock controller in communication with a drone remote control, the drone remote control in communication with the drone; The control terminal provides a meshed dock drone patrol inspection system, which is used to optimize the inspection time by obtaining patrol inspection targets corresponding to each dock based on the drone's current range and the distance from the patrol inspection targets to each dock, generate optimal patrol inspection routes for each drone based on the determined patrol inspection targets, and send them to the dock controller.

[0065] As shown in Figures 9(a) and 9(b), the dock described in this embodiment is a mobile drone dock including a mascon and a dock body, and the dock body includes a charging module 16, a drone parking space 17, an energy storage module 18 and a display module 19, of which, an attachment module is provided in the dock body, and the attachment module fixes the dock body with a screw-type automatic locking structure, and a drone fixing device that autonomously suppresses vibration in the horizontal and vertical directions is provided in the drone parking space, and the mascon is connected to the charging module and the attachment module respectively.

[0066] in particular, As shown in FIG. 10, the charging module includes several charging ports 20, a BMS (BATTERY MANAGEMENT SYSTEM) control board 21, a charging heat dissipation fan 22, a communication interface 23 and a charging indicator light 24. The drone parking space of the dock body can accommodate mainstream RTK (Real-time kinematic) drones on the market, and the drone parking space is provided with a fixing device for fixing the drone, in order to accommodate the mobile drone dock that is often used in different terrain environments and moves with a vehicle. As shown in Fig. 11, the fixing device includes a first clamping member 25 and a second clamping member 28, and the first clamping member and the second clamping member are connected by an elastic member 27 to form a clamping structure (similar to a clip structure).

[0067] Among them, the first clamping member 25 is fixed to the surface of the drone parking space, and as shown in Fig. 12, the second clamping member includes a handle 27-1, a first sleeve 27-2, a first spring 27-5, two telescopic rods 27-3 located at both ends of the sleeve, and a fixed end, the first spring is located in the middle of the sleeve, and both ends of the first spring are fixedly connected to one end of the two telescopic rods, respectively, the first spring applies a tensile force to the two telescopic rods toward the center of the sleeve, and the fixed end realizes the horizontal fixation of the drone, and at the same time, the fixed end realizes the vertical fixation of the drone based on the clamping structure formed in the fixing device. Based on the horizontal spring and vertical elastic material of the fixing device, the fixing device can be fixed well, and after being subjected to a situation such as shaking, the spring and elastic material act as a damper to absorb the horizontal and vertical forces of the drone, realizing vibration protection, and further ensuring the safety of the drone.

[0068] The dock body is provided with an attachment module for attaching the drone dock, of which the attachment module penetrates the dock body and is fixedly connected to the dock body. As shown in FIG. 13(a) to FIG. 13(c), the attachment module fixes the dock body with a screw-type automatic locking structure, and the screw-type automatic locking structure includes a second sleeve 28 and a double output shaft type motor 29 fixed to the center position of the sleeve. Both ends of the rotor of the double output shaft type motor are fixedly connected to a part of a screw 32, and the other end of the screw 32 is connected to one end of a spring slider 31 through a screw hole. The spring slider 31 moves horizontally and linearly with the rotation of the screw, and moves the second telescopic rod 30 fixedly connected to the other end of the spring slider to extend and retract. The spring slider 31 includes a first slider 31-1 and a second slider 31-3, and the first slider 31-1 and the second slider 31-3 are connected by a second spring 31-2. The first slider 31-1 of the spring slider 31 is provided with a screw hole that matches the screw 32, and the second slider 31-3 is provided with a circular hole at a position corresponding to the first slider 31-1, the diameter of the circular hole being larger than the outer diameter of the screw, and at the same time, a hole with a preset length is opened at one end of a fixed rod fixedly connected to the second slider, and the diameter of the hole is also larger than the outer diameter of the screw.

[0069] This is for the purpose of making the spring slider 31 and the second telescopic rod 30 easily telescopic in the horizontal direction within the sleeve.

[0070] In order to ensure automation of the installation process, a pressure sensor is provided on the first slider 31-1, and the pressure sensor is connected to the master controller. At the same time, the master controller is connected to the dual output shaft motor and controls the operation of the dual output shaft motor based on the comparison result between the acquired pressure value of the pressure sensor and a preset threshold value.

[0071] Specifically, the working mechanism of the mounting module is as follows. The double output shaft motor acts as a power core to rotate the screw, and the rotation of the screw displaces the power block along the horizontal direction, and the power block transmits thrust force through a spring to gradually bring the fixed end into contact with the cargo box (in this embodiment, it refers to the cabin of the pickup truck), and a pressure sensor is installed on the power block, and after the sensor receives that the reaction force from the cabin reaches a predetermined value, it forms a feedback and stops the rotation of the double output shaft motor to automatically lock. When the vehicle shakes, the spring acts as a damper to absorb the vibration and maintain the self-stable state of the mobile dock.

[0072] The drone performs work according to the patrol inspection task. The drone's autonomous patrol inspection software is deployed in the dock to perform detailed patrol inspection work based on a pre-prepared flight plan. The worker determines the drone's current status and specific work mode based on the indication status on the dock screen. After the drone completes the work, the worker manually replaces the drone's battery, fully exerting the worker's subjective initiative.

[0073] A display module is further connected to the master controller for displaying the status of the battery in the charging port of the charging module and issuing commands, where the issuing commands include issuing an installation command (i.e., installing the drone dock in the vehicle) and issuing a work task to the drone.

[0074] The energy storage module 18 is equipped with a dedicated charging gun to charge it as the dock's mobile work energy supply module. When the mobile drone dock accompanies the vehicle to the site to perform patrol inspection work, the energy storage module supports the power supply of each module in the dock, including the charging module, display module, master module, etc.

[0075] As shown in FIG. 14, after the mobile drone dock described in the present invention is installed in the pickup truck, the worker operates the mounting module through the display module to automatically lock it with the pickup truck. When the drone performs a patrol inspection task, the vehicle carries the mobile dock and arrives near the work site. After opening the dock, the worker turns on the drone fixing device to take out the drone, selects and installs the recommended battery in the charging module, and selects an appropriate patrol inspection route through the autonomous flight software in the dock. The drone autonomously completes the patrol inspection task. After the task is completed, the worker replaces the battery and returns the drone to the dock.

[0076] Example 3 As shown in FIG. 15 , the third embodiment of the present invention provides a drone patrol inspection method for a meshed dock, specifically including: Assuming that the drone's flight speed is V and the dock's position is the three-dimensional coordinate (0,0,0), this embodiment takes a tower or a pillar as an example of a patrol inspection target.

[0077] The position coordinates of several towers and pillars are (X1, Y1, Z1), (X2, Y2, Z2), ..., (Xn, Yn, Zn), respectively. The position of the dock is the center of a sphere or a plane, and is extended in three directions, X, Y, and Z axes. The coordinates of tower / pillar N are assumed to be (Xn, Yn, Zn). Since the straight line between the two points is the shortest, the straight-line distance that the drone flies from the dock to tower / pillar N is: The file is JPEG2024542562000002.jpg747.

[0078] The inspection complexity of a single tower / pole is determined according to the type of tower / pole (e.g., tension tower, straight tower, angle tower, etc.), and its patrol inspection complexity can be determined based on the 3D point cloud model of the tower / pole, where time Tn represents the complexity of tower / pole N, and its physical meaning is the time it takes for a drone to patrol and inspect the power tower / pole.

[0079] Take the following example: for a single power tower / pillar N, the time required to complete one patrol inspection is T=Sn / V+Tn+Sn / V, which includes three parts: the time Sn / V it takes for the drone to go from the dock to the tower / pillar N, the time Tn it takes to patrol the target object tower / pillar N, and the time Sn / V it takes for the drone to return to the dock after completing the patrol inspection.

[0080] What needs to be considered here is that if the drone's flight time is still sufficient after inspecting tower / pillar N, it can inspect two or even more towers / pillars in one flight task, in order to reduce the number of times the drone returns to the dock, that is, to reduce the number of times the drone inspects within the coverage area of ​​the dock, so as to achieve the goal of the optimal route and the shortest time.

[0081] In this embodiment, the complexity T of towers / pillars 1, 2, 3, ..., n and their three-dimensional coordinates (Xn, Yn, Zn) are known, and the towers / pillars closest to the dock origin are numbered 1, 2, 3, ..., n from the farthest tower / pillar to provide the basis for the subsequent planning method, the drone's cruising capability is T, the drone's patrol inspection speed is V, and the coordinates of the dock position are (0,0,0).

[0082] The method and steps for autonomously planning a route are as follows. The basic principle of route planning is to plan towers and columns far from the dock origin first, i.e. The process starts with the maximum value, JPEG2024542562000003.jpg747, and then decreases accordingly.

[0083] First, a task that can only inspect one tower is planned and confirmed. The satisfaction condition of this task is to screen the towers and columns that are far from the dock origin and have high tower / column complexity within the coverage area of ​​the dock, and execute it as a single tower task. The formula is as follows: Represented in JPEG2024542562000004.jpg1184, Among them, JPEG2024542562000005.jpg841, where the complexity of tower / pillar N is denoted as Tn in seconds, the three-dimensional coordinates of tower / pillar N are denoted as (Xn,Yn,Zn) in meters, the cruising capability of the drone is denoted as T in seconds, the patrol inspection speed of the drone is V in meters per second, and the position of the dock is (0,0,0) in meters.

[0084] The judgment principle of the above formula is as follows: After the drone finishes the patrol of a single tower / pillar, the remaining flight capacity is smaller than the complexity T of all other towers / pillars within the coverage area of ​​the dock, that is, this tower / pillar can be completed with only one patrol inspection task, and these towers / pillars are considered as the farthest route tasks in the coverage area of ​​the dock.

[0085] Next, some tower and pillar tasks JPEG2024542562000006.jpg876, but the surrounding towers and pillars n-1 However, the remaining range after inspecting the tower and column Tn is insufficient to complete one inspection, and the formula is as follows: JPEG2024542562000007.jpg769, and Represented in JPEG2024542562000008.jpg9146, Among them, T n-1 is the tower / pillar closest to Tn in a straight line, and its number is n-1, so it is closer to the dock origin, and at this point, the route planning for all single tower tasks has already been completed.

[0086] The next is a navigation task involving two towers / pillars, and the judgment principle is as follows: after the drone finishes patrolling a single tower / pillar, if the remaining flight capability can only continue to patrol one tower / pillar nearby the single tower / pillar, it will fly in a straight line to the nearby tower / pillar to continue patrolling, and then return in a straight line to the dock, and the navigation path will form a triangle.

[0087] Furthermore, the route planning sequence is from farthest to near, that is, from the highest number to the lowest. After the patrol inspection of a tower / pillar is completed, if there is still navigation capability, when searching for nearby towers / pillars, only towers / pillars with numbers lower than the current tower / pillar number can be searched, ensuring the clarity of the method.

[0088] The formula is expressed as follows: JPEG2024542562000009.jpg10156

[0089] The following is a route task that includes three towers and columns. The route task forms a rectangle, and at this time, it is necessary to search for nearby towers and columns in one direction. For example, the tower and column 3-2-1 on the left side of Figure 16 is a clockwise search method, which can avoid route overlap and optimize the route path. Because the distance to the dock origin of tower and column 3 is farther than that of tower and column 1, the route planning starts from tower and column 3. In fact, after the route task is confirmed, the patrol inspection can be performed in the reverse order. For example, the patrol inspection order of tower and column 3-2-1 or 1-2-3 is both possible, because the distance of the patrol inspection path is the same length. The reason why the route planned here is 3-2-1 is because it follows the tower and column planning principle from far to near, which calculates a shorter path than the reverse planning method from near to far.

[0090] By analogy, a route task including four or five towers / pillars can be sequentially planned and issued. For example, an irregular pentagon is planned for a route with four towers / pillars, and the sum of the lengths of the five sides is the total length of the route. As the number of route plans increases, it means that towers / pillars farther from the dock origin have already been planned in the route, and the closer the towers / pillars are to the dock origin, the larger the search range for nearby towers / pillars will be, because the closer the towers / pillars are to the dock origin, the larger their remaining sailing capacity will be.

[0091] As shown in Figure 16, the difference between the two figures is the route planning of tower-pillar 6, 1, 2, and 3. In the left figure, tower-pillar 6 is planned first, and then the tower-pillars in the nearby range are searched for based on the range capability. If the search range does not include tower-pillar 1, as shown in the left figure, tower-pillar 6 can only be planned for a single tower-pillar route task. Meanwhile, as shown in the right figure, if tower-pillar 1 is included in the planning range of tower-pillar 6, and the triangular route formed by tower-pillar 6, tower-pillar 1, and the dock origin meets the range capability of the drone through calculation, tower-pillar 6 and tower-pillar 1 will be planned for both tower-pillar routes.

[0092] In this embodiment, as shown in FIG. 17, meshing is realized by arranging multiple docks based on substations in an area where power towers and poles are evenly distributed, and arranging them in a mesh-like manner in the area. If there is a power tower or pole that needs to be inspected within the coverage area of ​​two docks, the method of generating the optimal route is as follows: First, determine which drone dock the power tower / pole is closest to, because the closer one has a shorter patrol inspection path. One tower / pole only needs to be included in one drone dock patrol inspection task once. Next, according to the above-mentioned single-dock route planning method, first number the power towers / pillars within the coverage area of ​​the dock in ascending order from the closest to the furthest, and the starting tower / pillar for route planning is the one with the highest number, i.e., the tower / pillar farthest from the drone dock, and judge whether the tower / pillar is a single-tower / pillar route task or a multi-tower / pillar route task from the furthest to the closest.

[0093] When a power tower / pole is included and covered by two drone docks or multiple drone docks, it is only necessary to determine which drone dock's route task the tower / pole belongs to.

[0094] The number of docks installed in a substation is not limited to one. For example, in order to improve the efficiency of patrol inspections, multiple docks can be installed in one substation, each completing patrol inspections of lines with different requirements, such as different directions or different voltage levels. It can be understood that such a meshed arrangement within a substation further enhances the significance and feasibility of meshing docks.

[0095] After all drone docks are bound to their corresponding towers and pillars and the route task planning is completed, each tower and pillar exists in only one route task of a certain drone dock, ensuring that there are no duplicated patrol inspection routes. The background control terminal derives the route that only this dock has based on the SN number of the drone dock, ensuring the one-to-one correspondence between each route and the drone dock.

[0096] Example 4 A fourth embodiment of the present invention provides a method for determining a task execution environment of a drone using the drone patrol inspection system of the meshed dock described in the first or second embodiment, the method comprising: Acquiring in-dock environment information and out-dock environment information within a sensing range; Select a target dock based on the drone position, determine corresponding flight influence factors based on a flight command, obtain corresponding flight environment data from external environment information of the selected target dock, determine flight conditions based on the flight environment data, and control the drone to return if the flight environment data does not satisfy the flight conditions; The method includes determining corresponding landing influence factors based on a return command, obtaining corresponding landing environment data and return environment data from the outside environment information and the inside environment information of the target dock based on the landing influence factors, controlling the landing method of the drone based on the landing environment data until the drone returns to the target dock, and adjusting the inside environment of the dock based on the return environment data.

[0097] In this embodiment, the in-dock environment information includes an in-dock temperature, an in-dock humidity, and an in-dock haze concentration; The dock environment information is collected by a temperature sensor, a humidity sensor, and a smoke sensor, Among them, the temperature sensor is used to collect the temperature of the internal environment of the dock. When the temperature is lower than the lower limit of the set temperature range, the heating function of the air conditioner is controlled to make the temperature inside the dock reach the normal operating range. When the temperature is higher than the upper limit of the set temperature range, the cooling function of the air conditioner is turned on to make the temperature inside the dock reach the normal operating range. The humidity sensor is used to detect the humidity of the internal environment of the dock. When the humidity inside the dock is higher than the set threshold, the moisture absorption function of the air conditioner is turned on. The fume sensor is used to detect the concentration of fume inside the dock.

[0098] In this embodiment, the outside-dock environment information includes wind speed, wind direction, outside-dock temperature, outside-dock humidity, rainfall, air pressure, solar radiation intensity, and visibility; The outside dock environment information is collected by a wind speed sensor, a wind direction sensor, a temperature sensor, a humidity sensor, a rain gauge, a barometer, a light sensor, and a visibility sensor; Among them, the wind speed sensor is used to measure the wind speed at the dock location, the wind direction sensor is used to measure the wind direction, the temperature sensor is used to measure the environmental temperature, the humidity sensor is used to measure the environmental humidity, the rain gauge is used to measure the amount of rain when it rains, and can be used to distinguish between light rain, medium rain, heavy rain, etc., the barometer is used to measure the local air pressure, the light sensor is used to measure the current solar radiation intensity, and the visibility sensor can continuously output the atmospheric visibility.

[0099] In this embodiment, the several sensors transmit collected data wirelessly.

[0100] As an optional embodiment, the wireless communication can adopt UWB wireless communication technology, which has characteristics such as low power consumption, high data transmission speed, strong interference resistance, and strong penetration ability.

[0101] The adoption of UWB wireless communication is only one possible embodiment shown in this embodiment, but it is not limited to such a wireless communication method, and it can be understood that in many embodiments, other wireless communication methods, such as 4G, 5G, etc., may be adopted according to the actual situation on site.

[0102] This embodiment merely lists and summarizes the types of data and sensors commonly used in the field, and it can be understood that in many more embodiments, sensor types may be added or removed depending on the actual situation.

[0103] In this embodiment, various sensors are used to collect data on the environment inside and outside the dock, and the collected sensor data is pre-processed. The pre-processing includes pre-processing the sensor data through a moving average low-pass filter to filter out hopping or abnormal environmental information, and obtaining relatively calm environmental information after pre-processing; In an alternative embodiment, the model of the moving average low pass filter is an output of an N-point moving average filter; The image is JPEG2024542562000010.jpg563.

[0104] In this embodiment, the influencing factors are classified according to the task command, and the flight conditions are judged according to the necessary influencing factors for different task commands; As shown in FIG. 18, specifically, the task command includes drone storage, drone charging, drone patrol inspection, dock self-check, dock opening and closing operation, dock open status, drone flight task, drone precision landing, drone alternative landing, etc. Specifically, the main influencing factors of drone storage, drone charging, and dock self-check are the dock environment information, including the dock temperature, dock humidity, and dock fume concentration, The main factors affecting drone patrol inspections include wind speed, wind direction, outside dock temperature, rainfall, barometer, solar radiation intensity, and visibility. The main factors influencing the opening and closing of the dock are the amount of rainfall and the concentration of smoke inside the dock. The main influencing factors in drone flight tasks are wind speed, wind direction, barometer and visibility. The main factors affecting precision landing of drones are wind speed, wind direction, solar radiation intensity, and visibility. The main factors influencing alternative drone landings are wind speed and direction.

[0105] As an alternative embodiment, taking the task of storing a drone as an example, the environmental impact factors under the conditions of the task are linked, and the conditions suitable for the task are judged by the threshold judgment method, as shown in FIG. Obtain the temperature, humidity and smoke concentration inside the dock, A temperature threshold, a humidity threshold, and a smoke threshold are preset; Determine whether the temperature inside the dock satisfies a temperature threshold condition. If the temperature inside the dock does not satisfy the temperature threshold condition, the temperature inside the dock is abnormal; If the temperature inside the dock satisfies the temperature threshold condition, determine whether the humidity inside the dock satisfies the humidity threshold condition. If the humidity inside the dock does not satisfy the humidity threshold condition, the humidity inside the dock is abnormal; If the humidity inside the dock meets the humidity threshold condition, then the smoke threshold is used to determine whether there is smoke inside the dock. If there is smoke inside the dock, then the smoke inside the dock is abnormal; if there is no smoke inside the dock, then the internal environment of the dock is normal and the drone can return to normal.

[0106] As an optional embodiment, the current judgment result and abnormality factors are packed into message information and pushed, and the output of the corresponding task is U8 type data, which represents the current judgment result and abnormality factors, of which 01 is the task number, and the latter 8 bits of data are used to represent the judgment result, and the judgment conclusion is the overall environmental judgment result, where 0 is abnormal and 1 is appropriate, and the rest are the judgment conclusions of the sensors in sequence, where 0 means that the current environmental items are abnormal, otherwise, the environment is appropriate, and if the environmental judgment result is 1, the judgment results of the sensors are all 1, otherwise, the data of the register at the sensor location is used to determine which current environment does not meet the needs of the current task, and by sequential inference, message information in different tasks is formed, and the judgment result can be directly called according to the current task status to determine whether to execute the task.

[0107] In an alternative embodiment, message information is sent outward at or above a set rate.

[0108] As an alternative embodiment, the method is applicable to a drone performing a flight task in a single dock and in the sensing range of the single dock, specifically: Obtaining internal environment information of a single dock and external environment information within a sensing range of the single dock; Determine corresponding flight influence factors according to the flight command, obtain corresponding flight environment data from the external environment information of the single dock, judge flight conditions based on the flight environment data, and control the drone to return if the flight environment data does not satisfy the flight conditions; The method includes determining corresponding landing influence factors based on a return command, obtaining corresponding landing environment data and return environment data from the external environment information and internal environment information of the single dock, controlling the landing method of the drone based on the landing environment data until the drone returns to the dock, and adjusting the internal environment of the dock based on the return environment data.

[0109] In this embodiment, a one-dock, one-drone approach is adopted, and all drone flight tasks are within the detection range of the dock, so the dock can collect environmental information during the drone's flight task and return flight in real time and make condition judgments.

[0110] As an optional embodiment, a method is adopted in which, if a drone flies out of the detection range of a dock, a plurality of docks arranged in a meshed manner and a drone performs a flight task within the detection range of the dock, and the specific method is as follows: Acquiring in-dock environment information and out-dock environment information within a sensing range; Select a target dock based on the drone position, determine corresponding flight influence factors based on a flight command, obtain corresponding flight environment data from the outside environment information of the dock, determine flight conditions based on the flight environment data, and control the drone to return if the flight environment data does not satisfy the flight conditions; The method includes determining corresponding landing influence factors based on a return command, obtaining corresponding landing environment data and return environment data from the off-dock environment information and on-dock environment information of the target dock, controlling the landing method of the drone based on the landing environment data until the drone returns to the target dock, and adjusting the on-dock environment based on the return environment data.

[0111] In this embodiment, the distance between the docks does not exceed the detection distance, that is, when the drone leaves the detection range of one of the docks, it will enter the detection range of the other dock. Therefore, it is determined whether the drone is within the detection range of the dock based on the drone's position, and the dock that is within the detection range is set as the target dock, and environmental information during the drone's flight task and return flight is collected according to the target dock; If the detection ranges of two docks overlap, the dock with the closest distance will be selected as the target dock based on the distance between the drone and the dock.

[0112] In this embodiment, the flow of judging the flight conditions of the drone in the above method is specifically as follows: receiving sensor information and task instructions; In response to a task command, determine the classification of the task and perform task decomposition; If the command is a patrol inspection, determine whether the external environment of the current dock is suitable for the patrol inspection task, and if not, terminate the task and upload the cause of the task termination; If it is suitable to perform the task, perform a self-check of the dock; After passing the dock self-check, the drone will take off and perform patrol inspection tasks; During the execution of the patrol inspection task, determining whether an unfavorable condition for the patrol inspection has appeared in the external environment; If the external environment is not suitable for flight or a return flight command is received by the drone, the drone returns and determines whether the current environment meets the precision landing conditions; If the precision landing is met, perform precision landing of the drone, and determine whether the dock meets the requirements for storing and charging the drone. If the internal environment of the dock is abnormal, adjust the internal environment of the dock until the drone can be charged and stored; If precision landing is not met, perform an alternative landing of the drone; If the alternative landing of the drone is not met, the drone is forcibly landed and the forced landing conditions and adverse factors are uploaded.

[0113] In this embodiment, the process of adjusting the internal environment of the dock includes: If the temperature inside the dock is lower than the lower limit of the set temperature range, control the heating function of the air conditioner to make the temperature inside the dock reach the normal operating range; If the temperature inside the dock is higher than the upper limit of the set temperature range, turn on the air conditioner's cooling function to allow the temperature inside the dock to reach the normal operating range; If the humidity inside the dock is higher than the set threshold, turn on the air conditioner's moisture absorption function, If the in-dock fumes are abnormal, performing an alternative landing of the drone.

[0114] Example 5 As shown in FIG. 20, the fifth embodiment of the present invention utilizes the drone patrol inspection system of the meshed dock described in the first or second embodiment, Obtaining drone positioning data; Based on the acquired positioning data, determine whether the drone is located within a preset landing range, and if the drone is not located within the preset landing range, control the drone to move until the position request is satisfied; After it is determined that the drone is located within the preset landing range, when the drone is located at a position that is a first preset distance away from the landing point, obtain image data or video data below the drone, and if the precision landing range code cannot be identified based on the obtained image data or video data, control the drone to descend to a position that is a third preset distance away from the landing point and identify the precision landing range code again until the precision landing range code is identified; The present invention provides a method for controlling precision landing of a drone, the method including: when a precision landing range code is identified based on the acquired image data or video data, controlling the drone to descend to a position a second preset distance away from the landing point and to again acquire image data or video data below the drone; and when a precision landing position code is identified again based on the acquired image data or video data, controlling the drone to descend to a position a fourth preset distance away from the landing point and to land the drone.

[0115] In this embodiment, the precision landing range code and the precision landing position code are one big one and one small one next to each other, the big one is called the precision landing range code, which is used in high altitude, mainly used to determine the rough position where the drone lands, and constantly land and adjust the position and attitude; the small one is called the precision landing position code, which is used in low altitude, the drone starts to identify, constantly adjust the position and attitude, and finally lands on the precision landing position code of the small drone.

[0116] The method described in this embodiment firstly utilizes the RTK technology of the drone, so that the drone can quickly and accurately return above the landing point after completing the task. The RTK technology can reduce the error of the drone's flight to the centimeter level. In this way, the drone does not need to update its coordinates multiple times, and can quickly return to the landing point. After reaching the landing range, it starts imaging to search for the landing range code, and completes the identification within 0.7 seconds after completing the image reception, then returns to adjust the drone's appearance, and descends to a height of 20 cm to achieve instrument landing, thereby realizing the full automation of the drone's patrol inspection task.

[0117] Example 6 Example 6 of the present invention provides a drone patrol inspection method based on visual movement tracking, using the drone patrol inspection system of the meshed dock described in Example 1 or Example 2, wherein: The drone is equipped with a three-axis head, an RTK positioning module and a front-end AI processing module. A camera and a video camera are attached to the three-axis head. The camera is a monocular zoomable camera, and the video camera is used to obtain video information of the tower / pillar, wherein the camera and the video camera are integrated into one lens.

[0118] The RTK positioning module is used to measure the drone's three-dimensional coordinate information. The front-end AI processing module fits the drone's flight control data, the data of the RTK positioning module and the images collected by the zoom camera, issues flight control commands to control the drone's flight, controls the camera head to adjust the camera angle and zooming, locks on the patrol inspection target and takes photos, uses the visual zoom wide-angle camera to take photos during the flight process approaching the hovering point, calculates the coordinate values ​​(GPS values) of the photos taken and the attitude of the camera head, identifies the patrol inspection target in the photo according to the camera imaging principle, adjusts the position of the drone head according to the current GPS position and three-dimensional speed of the drone, as well as the roll angle, pitch angle and yaw angle of the attitude of the camera head, and adjusts the position of the drone head according to the Kalman filter algorithm, makes the zoom camera lock on the target inspection point of the tower / pillar by zooming, and finally takes photos to complete the information collection of the target inspection point of the tower / pillar, thereby improving the accuracy of the information collection of the patrol inspection target and the quality of the collected images.

[0119] During the flight process from when the drone enters the patrol inspection target to when it leaves the patrol inspection target, the drone always flies along the set flight path, and uses the Kalman filter algorithm to fit the current position and speed data to adjust the pan head attitude and camera zooming in real time, thereby realizing camera movement tracking and locked-on shooting of the patrol inspection target.

[0120] When using a method based on visual motion tracking to control the rotation of the head, the head has m degrees of freedom, and the angular velocity of the head rotation is JPEG2024542562000011.jpg629, and the linear velocity at the end is JPEG2024542562000012.jpg630, and the two have the following relationship: JPEG2024542562000013.jpg622 Among them, The image is JPEG2024542562000014.jpg3643.

[0121] Rotation matrix that transforms from the geodetic coordinate system to the camera coordinate system Calculate JPEG2024542562000015.jpg77, JPEG2024542562000016.jpg31144 Among them, the subscript cw represents the abbreviation for transforming from the geodetic coordinate system to the camera coordinate system, JPEG2024542562000017.jpg639 represents the matrices that need to be rotated around the x, y and z axes to get from the camera coordinate system to the geodetic coordinate system, JPEG2024542562000018.jpg1135 is the roll angle, pitch angle, and yaw angle of the camera head posture, respectively. Based on the initial orientation of the camera, It is necessary to multiply it by JPEG2024542562000019.jpg69, and at this time, JPEG2024542562000020.jpg880. During the ceremony, The image is JPEG2024542562000021.jpg1931.

[0122] As shown in FIG. 21, the specific steps include:

[0123] In S1, in response to a patrol inspection request, before the drone enters the hovering point at a constant speed, a monocular zoom (telephoto mode) camera on the pan head is used to obtain a real-time wide-angle image of the patrol inspection target, and then proceeds to the next step.

[0124] In S2, it is determined whether the patrol inspection target is located in the captured real-time image. If the answer is YES, proceed to S4. If the answer is NO, control the camera head attitude in an "O" shape to search for the patrol inspection target in the real-time image. After the patrol inspection target is found, proceed to the next step.

[0125] In S3, the front-end AI processing module adopts the Kalman filter algorithm based on information such as the patrol inspection target position in the real-time image, the drone shooting position, and the three-axis head attitude, to fit and output the drone position, head attitude position, and camera focal length mode that need to be adjusted, and then transfers to the execution of S2.

[0126] In S4, before the drone reaches the hovering point at a constant speed, the monocular zoom (close focal length mode) camera on the pan head is adjusted, and the front-end AI processing module reversely adjusts the attitude of the three-axis pan head in real time according to the three-dimensional direction in which the drone flies at a constant speed, so as to achieve that the patrol inspection target is locked on to the center position of the real-time image of the monocular zoom camera, and then proceeds to the next step.

[0127] In S5, the drone reaches the position of the hovering point, i.e., reaches the front direction just in front of the patrol inspection target, confirms that the inspection viewpoint is locked on to the center position of the real-time image of the monocular zoom camera, takes a photo, and proceeds to the next step.

[0128] In S6, the camera captures the image, the front-end AI processing module processes the photos, and controls the drone to perform the task of the next hovering point until it has captured the images of all the hovering points and safely returns, and then re-executes S1.

[0129] Furthermore, the specific process of S2 is as follows: The Faster-RCNN algorithm is adopted to input the picture into the CNN to perform feature extraction, and then determine whether the patrol inspection target exists in the picture.

[0130] Furthermore, the specific process of S3 is as follows: This step assumes that the presence of a patrol target object in the image has already been identified by S2; In S3.1, the rotation direction of the camera platform is determined based on the position of the target object at the inspection viewpoint in the image, which is the direction to offset the tower / pillar to the center of the image. The camera platform is rotated by the minimum unit first, the tower / pillar image at the current position is obtained, and its features are extracted. In S3.2, the features of the two pictures are matched, and the offset of the matching point at the pixel point is calculated. In S3.3, the rotation amount of the pan head is obtained based on the linear mapping relationship between the feature offset amount and the rotation amount of the pan head; In S3.4, adjust the head position according to the amount of rotation, and re-execute S2.

[0131] Furthermore, the specific process of S4 is as follows: Assuming that the initial state is already set to the center of the camera image as described in step 3, In S4.1, the current drone position and the three-dimensional vector direction P in which the drone is moving are calculated using the RTK and accelerometer of the drone. In S4.2, adjust the camera's motion vector on the platform so that it is exactly equal to the magnitude and opposite to the direction of the drone's motion vector. In S4.3, calculate the offset amount of the pixel of the central target of the camera at the current time according to the method described in S4.2, if there is no offset amount, it is considered that the moving tracking of the camera of the pan head to the patrol inspection target object is relatively stationary; if not, proceed to the next step; In S4.4, the rotation amount of the pan-and-tilt head is obtained based on the linear mapping relationship between the feature offset amount of the pixel in the center of the image and the rotation amount of the pan-and-tilt head, and then the camera of the pan-and-tilt head is fine-tuned to re-lock on the target object of the inspection viewpoint to the center of the camera of the pan-and-tilt head.

[0132] The above is only a preferred embodiment of the present invention, and does not limit the present invention. Those skilled in the art can make various modifications and changes to the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention should be included within the protection scope of the present invention. [Explanation of symbols]

[0133] 1 Tower column, 2 Dock bottom support, 3 Dock, 4 Landing platform, 5 Top cover, 6 Dock body, 7 Rotating rod, 8 Second motor, 9 First motor, 10 Second centering rod, 11 Charging rod, 12 Charging port, 13 First centering rod, 14 Rack, 15 Fixed base, 16 Charging module, 17 Drone parking space, 18 Energy storage module, 19 Display module, 20 Charging port, 21 BMS control board, 22 ··Heat dissipation fan, 23···Communication interface, 24···Charging indicator light, 25···First clamping member, 26···Elastic member, 27···Second clamping member, 27-1···Handle, 27-2···First sleeve, 27-3···First telescopic rod, 27-4···Fixed end, 27-5···First spring, 28···Second sleeve, 29···Dual output shaft motor, 30···Second telescopic rod, 31···Spring slider, 31-1···First slider, 31-2···Second spring, 31-3···Second slider, 32···Screw, 33···Drone, 34···Dock.

Claims

1. a plurality of docks arranged in a meshed configuration, each dock being used to house at least one drone; The dock includes a dock controller in communication with the control terminal, the dock controller in communication with a drone remote control, the drone remote control in communication with the drone; The control terminal is used to optimize the inspection time by obtaining the inspection target corresponding to each dock based on the current range of the drone and the distance from the inspection target to each dock, and generate an optimal inspection route for each drone based on the determined inspection target and send it to the dock controller; A drone patrol inspection system for a meshed dock.

2. The dock is The present invention relates to a dock body, and a mounting mechanism, a vertical fixing mechanism and a horizontal fixing mechanism provided in the dock body, the mounting mechanism including an extendable landing platform and a first motor, the landing platform being driven by the first motor; The vertical fixing mechanism includes a first centering rod having one end provided on a side wall of the dock body via a pivot shaft and a gear, and the landing platform is provided with a rack that engages with the gear, and the first centering rod is driven to rotate around the pivot shaft by the engagement between the gear and the rack; The lateral fixing mechanism includes a rotating rod, both ends of which are provided on the side walls of the dock body, a second centering rod provided on the rotating rod, and a second motor, the rotating rod being driven by the second motor to rotate along a direction opposite to the moving direction of the landing platform relative to the dock body, and driving the second centering rod to move along a direction perpendicular to the moving direction of the landing platform. The meshed dock drone patrol inspection system according to claim 1 .

3. When the gear and the rack mesh, the first centering rods provided on both side walls rotate around their axes, causing the other ends of the two first centering rods to move to intermediate positions or to spread apart in both directions. The meshed dock drone patrol inspection system according to claim 2 .

4. A second centering rod is provided at each end of the pivot rod. When the landing platform is driven and reset, the pivot rod rotates in the positive direction, and the second centering rods at both ends move along the pivot rod to an intermediate position to laterally restrain the drone; By meshing the rack with the gear, the first centering rods on both sides can be moved to a middle position to restrain the drone vertically. The meshed dock drone patrol inspection system according to claim 3.

5. A second centering rod is provided at each end of the pivot rod. When the landing platform is pushed out from the dock body, the pivot rod rotates in the opposite direction, and the second centering rods at both ends move to both sides along the pivot rod to release the lateral constraint on the drone; By meshing the rack with the gear, the first centering rods on both side walls can be moved to spread out to both sides to release the vertical restraint on the drone. The meshed dock drone patrol inspection system according to claim 3.

6. The dock includes a dock body having a drone parking space, a charging module, and an energy storage module therein; The dock body is provided with an attachment module that fixes the dock body with a screw-type automatic locking structure, the drone parking space is provided with a drone fixing device that autonomously suppresses vibrations in the horizontal and vertical directions, and the dock controller communicates with the charging module and the attachment module, respectively. The meshed dock drone patrol inspection system according to claim 1 .

7. The screw-type automatic locking mechanism includes a dual output shaft motor fixed to the center of the sleeve and the second sleeve, a screw is connected to each end of the rotor of the dual output shaft motor, and the other end of the screw is connected to one end of a spring slider through a screw hole, and the spring slider moves linearly with the rotation of the screw, and moves the telescopic rod fixedly connected to the other end of the spring slider to telescope. The meshed dock drone patrol inspection system according to claim 6.

8. The drone is equipped with a three-axis head, an RTK positioning module and a front-end AI processing module. A camera and a video camera are attached to the three-axis pan head, the camera is a monocular camera capable of zooming, the video camera is used to obtain video information of the tower / pillar, the camera and the video camera are integrated into one lens, The RTK positioning module is used to measure the drone's three-dimensional coordinate information. The front-end AI processing module is used to fit the drone's flight control data, the data of the RTK positioning module and the images collected by the zoom camera, issue flight control commands to control the drone's flight, control the pan head to adjust the camera angle and zoom, lock on the patrol inspection target and take photos; if the patrol inspection target is not located in the center position of the camera image, a method based on visual movement tracking is used to control the rotation of the pan head, and determine the rotation direction of the pan head according to the position of the patrol inspection target in the image; The meshed dock drone patrol inspection system according to claim 1 .

9. Obtaining a distance from a patrol inspection target to each dock; selecting a dock that is closest to the patrol inspection target as the optimal dock; Sequentially determining each patrol inspection target to obtain a patrol inspection target corresponding to each dock; Each dock performs a planning task for a patrol inspection task, the task comprising: The numbering of the patrol inspection targets is based on the distance from the patrol inspection target to the dock within the dock range, and the numbering rule is that the farther the distance from the patrol inspection target to the dock, the larger the number; For each patrol inspection target within the dock range, determine whether the difference between the total flight time of the drone and the time required for the patrol inspection target to perform one patrol inspection alone is smaller than the minimum time required for the other patrol inspection targets within the dock range to perform one patrol inspection alone, and if the result is YES, set this patrol inspection target as a single tower task; If NO, set this patrol inspection target as the current patrol inspection target, and determine whether the sum of the time from the dock to the current patrol inspection target, the patrol inspection time of the current patrol inspection target, the patrol inspection time from the current patrol inspection target to the nearest secondary patrol inspection target whose number is smaller than that of the current patrol inspection target, the time from the current patrol inspection target to the secondary patrol inspection target, and the time from the secondary patrol inspection target to the dock is greater than the total flight time of the drone, if YES, set the current patrol inspection target as a single tower task, if NO, execute the route task of two towers / pillars, and perform patrol inspections of the current patrol inspection target and the secondary patrol inspection target in sequence. A method for drone patrol inspection of a meshed dock, comprising:

10. Determining a comprehensive patrol inspection time of the next node, and if the comprehensive patrol inspection time is greater than the total flight time of the drone, only performing the current patrol inspection task; otherwise, continuing to perform the determination of the comprehensive patrol inspection time of the next node; The method for drone patrol inspection of a meshed dock as claimed in claim 9.

11. The meshed dock drone patrol inspection system according to any one of claims 1 to 8 is applied to the meshed dock drone patrol inspection system, Acquiring in-dock environment information and out-dock environment information within a sensing range; Select a target dock based on the drone position, determine corresponding flight influence factors based on the flight command, obtain corresponding flight environment data from the external environment information of the selected target dock, and control the drone to return if the flight environment data does not satisfy the flight conditions; Determining corresponding landing influence factors according to a return command, obtaining corresponding landing environment data and return environment data from the outside environment information and the inside environment information of the target dock according to the landing influence factors, controlling the landing manner of the drone according to the landing environment data until the drone returns to the target dock, and adjusting the inside environment of the target dock according to the return environment data; Selecting the target dock based on the drone position specifically includes: determining a detection range of the dock where the drone is located based on the drone position, and determining the dock within the detection range as the target dock; and if two or more docks are within the detection range, determining the dock closest to the drone as the target dock. A method for determining a task execution environment for a drone, comprising:

12. The dock environment information includes a temperature inside the dock, a humidity inside the dock, and a smoke concentration inside the dock, The outside-dock environment information includes wind speed, wind direction, outside-dock temperature, outside-dock humidity, rainfall, air pressure, solar radiation intensity, and visibility. The method for determining a task execution environment of a drone as described in claim 11.

13. In the process of determining corresponding flight affecting factors based on flight commands, the flight commands include drone patrol inspection, dock opening and closing operations, and drone flight tasks; The flight influence factors corresponding to drone patrol inspections include wind speed, wind direction, outside dock temperature, rainfall, air pressure, solar radiation intensity and visibility. Flight impact factors corresponding to the opening and closing of the dock include precipitation and haze concentration inside the dock; Flight influence factors corresponding to drone flight tasks include wind speed, wind direction, air pressure and visibility; The method for determining a task execution environment of a drone as described in claim 12.

14. In the process of determining corresponding landing affecting factors based on the return command, the commands affected by the return command include drone storage, drone charging, dock self-check, drone precision landing, and drone alternative landing; The landing influence factors corresponding to drone storage, drone charging and dock self-check include the temperature in the dock, the humidity in the dock and the haze concentration in the dock; The landing influence factors for precision landing of drones include wind speed, wind direction, solar radiation intensity and visibility. Landing influence factors that support alternative drone landings include wind speed and direction, The method for determining a task execution environment of a drone as described in claim 12.

15. If the current flight command is a drone patrol inspection command and the current flight environment data meets the flight conditions, performing a dock self-check; if the dock self-check is passed, controlling the drone to take off and perform a patrol inspection task; and continuously detecting whether the flight environment data meets the flight conditions during the process of performing the drone patrol inspection task. The method for determining a task execution environment of a drone as described in claim 11.

16. The method further includes: determining whether the current landing environment data meets the precision landing conditions; if the precision landing conditions are met, performing a precision landing of the drone; determining whether the dock meets the requirements for storing and charging the drone; if the environment inside the dock is abnormal, adjusting the environment inside the dock; if the precision landing conditions are not met, controlling the drone to make an alternative landing; and if the alternative landing of the drone is not possible, controlling the drone to make a forcible landing. The method for determining a task execution environment of a drone as described in claim 15.

17. The meshed dock drone patrol inspection system according to any one of claims 1 to 8 is applied to the meshed dock drone patrol inspection system, Obtaining drone positioning data; Based on the acquired positioning data, determine whether the drone is located within a preset landing range, and if the drone is not located within the preset landing range, control the drone to move until the position request is satisfied; After it is determined that the drone is located within the preset landing range, when the drone is located at a position away from the landing point by a first preset distance, obtain image data or video data below the drone, and if the precision landing range code cannot be identified based on the obtained image data or video data, control the drone to descend to a position away from the landing point by a third preset distance until the precision landing range code is identified, and to identify the precision landing range code again; if a precision landing range code is identified based on the acquired image data or video data, controlling the drone to descend to a position a second preset distance away from the landing point and again acquire image data or video data below the drone, and if a precision landing position code is identified based on the again acquired image data or video data, controlling the drone to descend to a position a fourth preset distance away from the landing point and controlling the drone to land. A method for controlling precision landing of a drone.

18. The meshed dock drone patrol inspection system according to claim 8 is applied to the meshed dock drone patrol inspection system, S1: before the drone enters the detection point at a constant speed in response to a patrol inspection request, the image acquisition module on the camera platform is used to acquire a real-time wide-angle image of the patrol inspection target; It is determined whether the patrol inspection target is located in the captured real-time image. If the result is YES, the process proceeds to step S3. If the result is NO, the process controls the movement of the camera platform to change its attitude until the patrol inspection target is found in the real-time image. The processing module adopts a Kalman filter algorithm to fit the drone's shooting position and the head attitude based on the information of the patrol inspection target position, the drone's shooting position, and the head attitude in the real-time image, and determines the focal length mode of the image acquisition module S3; Control the drone to fly at a constant speed to the shooting position obtained by fitting. During the flight, the processing module adjusts the attitude of the pan head in real time inversely according to the three-dimensional direction in which the drone flies at a constant speed, and adjusts the focal length mode of the image acquisition module S4, so as to achieve that the patrol inspection target is locked on to the setting area of ​​the real-time image of the image acquisition module; The drone arrives at the shooting position, confirms that the patrol inspection target position is within the setting area of ​​the real-time image of the image collection module, locks on the inspection viewpoint and collects images S5; The processing module processes the captured picture, and controls the drone to perform the task of the next detection point, and re-executes S1 until completing the image capture task of all detection points. A drone patrol inspection method based on visual movement tracking, comprising:

19. In S2, The specific process of determining whether the inspection target is located in the captured real-time image includes: inputting the image into the CNN using the Faster-RCNN algorithm, performing feature extraction, and then determining whether the inspection target exists in the image; The drone patrol inspection method based on visual movement tracking as claimed in claim 18.

20. Specifically, S3 is S3.1: determining the rotation direction of the camera platform, which is the direction to offset the tower / pillar to the center of the image, based on the position of the target object of the inspection viewpoint in the image, and first rotating the camera platform by the minimum unit, acquiring the tower / pillar image at the current position, and extracting its features; S3.2: Matching the features of the two tower / pillar images and calculating the offset amount at the pixel point of the matching point in the image; S3.3 obtains a rotation amount of the pan head based on a linear mapping relationship between the feature offset amount and the rotation amount of the pan head; S3.4 adjusts the position of the camera head according to the amount of rotation. The drone patrol inspection method based on visual movement tracking as claimed in claim 18.

21. In S4, In S4.1, the current drone position and the three-dimensional vector direction P in which the drone is about to move are calculated based on the drone position information and acceleration information; In S4.2, adjust the motion vector of the image acquisition module of the camera head so that the magnitude of the motion vector of the drone is equal to and opposite to the direction of the motion vector of the drone; In step S4.3, calculate the offset amount of the pixel of the central target of the image acquisition module at the current time. If there is no offset amount, determine that the moving tracking of the image acquisition module of the pan head to the patrol inspection target object is in a relatively stationary state; if there is an offset amount, proceed to step S4.4; In S4.4, obtain the rotation amount of the pan head according to the linear mapping relationship between the feature offset amount of the pixel in the center of the image and the rotation amount of the pan head, and then fine-tune the position of the image acquisition module of the pan head to re-lock on the target object of the inspection point to the center of the image acquisition module of the pan head; The drone patrol inspection method based on visual movement tracking as claimed in claim 18.

Citation Information

Patent Citations

  • Path optimization method for multi-unmanned aerial vehicle cooperative inspection task

    CN111220159A

  • Multi-machine cooperative inspection system

    CN112596543A

  • Unmanned aerial vehicle carrying, flying and recycling system for square cabin

    CN112722307A

  • Unmanned aerial vehicle uninterrupted inspection system

    CN214474627U

  • Method and system for controlling the flight of an unmanned aerial vehicle

    JP2018537335A