Underground unmanned aerial vehicle obstacle avoidance method, computer equipment and computer readable storage medium

Abstract

The invention discloses an underground unmanned aerial vehicle obstacle avoidance method, computer equipment and a computer readable storage medium, and relates to the technical field of unmanned aerial vehicles. The method comprises the following steps: acquiring initial point cloud data corresponding to an underground environment, and preprocessing the initial point cloud data to obtain preprocessed point cloud data; constructing an underground three-dimensional map according to the preprocessed point cloud data based on a synchronous positioning and map construction algorithm; non-ground point clouds are determined according to the underground three-dimensional map, and obstacle features are extracted from the non-ground point clouds; determining a target obstacle avoidance area according to the unmanned aerial vehicle positioning data and the remote controller instruction data; and performing unmanned aerial vehicle obstacle avoidance according to the target obstacle avoidance area and the obstacle characteristics. Thus, the point cloud data in the underground environment is obtained through the laser radar, the obstacle features are extracted from the point cloud data, obstacle avoidance is performed by combining the obstacle features and the target obstacle avoidance area, and various obstacles can be effectively recognized and avoided even in the dark underground environment.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles, and in particular to an underground unmanned aerial vehicle obstacle avoidance method, a computer device and a computer-readable storage medium. Background Art

[0002] The complex and ever-changing underground environment is usually full of unpredictable obstacles, such as protruding rocks, crisscrossing pipes, hanging cables, and possible waterlogging or collapsed areas, which pose a great threat to the flight safety of drones. At present, traditional visual obstacle avoidance methods often lose their effectiveness due to lack of light and cannot effectively identify and avoid various obstacles. Summary of the invention

[0003] In view of this, the purpose of the present invention is to overcome the shortcomings of the prior art and provide an underground UAV obstacle avoidance method, computer equipment and computer-readable storage medium, which are used to use point cloud data obtained by lidar to detect whether there are obstacles in a set area, so as to determine whether the UAV should brake to avoid the obstacle.

[0004] The present invention provides the following technical solutions: In a first aspect, the present invention provides an underground UAV obstacle avoidance method, comprising: Acquiring initial point cloud data corresponding to the underground environment, and preprocessing the initial point cloud data to obtain preprocessed point cloud data; Based on the synchronous positioning and mapping algorithm, construct a three-dimensional map of the well according to the pre-processed point cloud data; Determine a non-ground point cloud according to the underground three-dimensional map, and extract obstacle features from the non-ground point cloud; Determine the target obstacle avoidance area based on the drone positioning data and remote controller command data; The drone avoids obstacles according to the target obstacle avoidance area and the obstacle characteristics.

[0005] In one embodiment, the drone positioning data includes current position data and current azimuth, and the remote controller command data includes the drone expected speed, the drone horizontal expected obstacle avoidance distance, the drone vertical expected obstacle avoidance distance, and the drone maximum obstacle avoidance distance; Determining the target obstacle avoidance area according to the drone positioning data and the remote controller command data includes: Based on the size of the component data of the expected speed of the drone, the target obstacle avoidance area is determined according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas; The short-distance horizontal obstacle avoidance area is calculated according to the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone; Calculate the long-distance horizontal obstacle avoidance area according to the current position data, the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance; Calculate the vertical obstacle avoidance area according to the current position data, the drone body data and the vertical expected obstacle avoidance distance of the drone; The obstacle avoidance areas in the horizontal and vertical directions are calculated according to the current position data, the current azimuth angle, the expected speed of the drone and the expected horizontal obstacle avoidance distance of the drone.

[0006] In one embodiment, the desired speed of the drone includes an x ​​speed component, a y speed component, and a z speed component, and the target obstacle avoidance area is determined based on the size of the component data of the desired speed of the drone according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas, including: If the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is within a preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is within the preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the x velocity component is within the preset range, the absolute value of the y velocity component is within the preset range, and the absolute value of the z velocity component is greater than 0, the vertical obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas; If the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas.

[0007] In one embodiment, the calculating the short-distance horizontal obstacle avoidance area according to the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone includes: Calculate and dynamically adjust the safety distance according to the expected obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone; Determine a preset sector angle according to the tangent function and the desired speed of the drone; Determine the vertical spatial range of the drone according to the drone fuselage data; The short-range horizontal obstacle avoidance area is determined according to the current position data, the dynamically adjusted safety distance, the preset fan-shaped body angle and the vertical spatial range of the UAV.

[0008] In one embodiment, the calculating the long-distance horizontal obstacle avoidance area according to the current position data, the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance includes: Based on the sine function and the cosine function, determining the obstacle avoidance range rectangular boundary and the rotation matrix according to the current azimuth angle, the drone body data, the horizontal expected obstacle avoidance distance of the drone and the maximum obstacle avoidance distance of the drone; The long-distance horizontal obstacle avoidance area is calculated according to the current position data, the obstacle avoidance range rectangular boundary and the rotation matrix.

[0009] In one embodiment, the calculating the vertical obstacle avoidance area according to the current position data, the drone body data and the vertical expected obstacle avoidance distance of the drone includes: Determine a preset cube range according to the drone fuselage data and the vertical expected obstacle avoidance distance of the drone; The vertical obstacle avoidance area is calculated according to the current position data and the preset cube range.

[0010] In one embodiment, the calculating the obstacle avoidance areas in the horizontal and vertical directions according to the current position data, the current azimuth, the expected speed of the drone, and the expected horizontal obstacle avoidance distance of the drone includes: Based on the sine function and the cosine function, the preset sphere local area range is determined according to the dynamically adjusted safety distance and the preset sector angle; The obstacle avoidance areas in the horizontal and vertical directions are calculated according to the current position data and the preset local area range of the sphere.

[0011] In one embodiment, the preprocessing of the initial point cloud data to obtain preprocessed point cloud data includes: De-noising the initial point cloud data based on a statistical filtering algorithm to obtain first point cloud data; The first point cloud data is filtered based on a voxel filtering algorithm to obtain the preprocessed point cloud data.

[0012] In a second aspect, the present invention proposes a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the underground UAV obstacle avoidance method as described in the first aspect is implemented.

[0013] In a third aspect, the present invention proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the underground UAV obstacle avoidance method as described in the first aspect.

[0014] The underground unmanned aerial vehicle obstacle avoidance method, computer device and computer-readable storage medium disclosed in the present invention obtains initial point cloud data corresponding to the underground environment, and pre-processes the initial point cloud data to obtain pre-processed point cloud data; constructs an underground three-dimensional map based on the pre-processed point cloud data based on the synchronous positioning and map construction algorithm; determines the non-ground point cloud according to the underground three-dimensional map, and extracts obstacle features from the non-ground point cloud; determines the target obstacle avoidance area according to the unmanned aerial vehicle positioning data and the remote control command data; and performs unmanned aerial vehicle obstacle avoidance according to the target obstacle avoidance area and the obstacle features. In this way, the point cloud data in the underground environment is obtained by laser radar, and obstacle features are extracted therefrom; at the same time, the target obstacle avoidance area is determined according to the unmanned aerial vehicle positioning data and the remote control command data, so as to combine the obstacle features and the target obstacle avoidance area to perform obstacle avoidance, and then, even in a dark underground environment, various obstacles can be effectively identified and avoided, ensuring the safety and smoothness of the unmanned aerial vehicle's movement. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present invention, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope of protection of the present invention. In each of the drawings, similar components are numbered similarly.

[0016] Figure 1 A first flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 2 A second flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 3 A third flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 4 A fourth flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 5 A fifth flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 6A sixth flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 7 The seventh flow chart of the underground UAV obstacle avoidance method proposed in this embodiment is shown; Figure 8 A schematic diagram of the target obstacle avoidance area proposed in this embodiment is shown; Fig. 9 A structural schematic diagram of the underground UAV obstacle avoidance device proposed in this embodiment is shown.

[0017] Description of the accompanying drawings: 900 - underground UAV obstacle avoidance device; 901 - acquisition module; 902 - construction module; 903 - extraction module; 904 - determination module; 905 - obstacle avoidance module. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0019] The components of the embodiments of the present invention generally described and shown in the drawings herein may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0020] Hereinafter, the terms "including", "having" and their cognates, which may be used in various embodiments of the present invention, are intended only to indicate specific features, numbers, steps, operations, elements, components or combinations of the foregoing items, and should not be understood as first excluding the existence of one or more other features, numbers, steps, operations, elements, components or combinations of the foregoing items or adding the possibility of one or more features, numbers, steps, operations, elements, components or combinations of the foregoing items.

[0021] Furthermore, the terms “first”, “second”, “third”, etc. are merely used for distinguishing descriptions and are not to be understood as indicating or implying relative importance.

[0022] Unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meanings as those generally understood by those skilled in the art to which the various embodiments of the present invention belong. The terms (such as those defined in generally used dictionaries) will be interpreted as having the same meanings as the contextual meanings in the relevant technical field and will not be interpreted as having idealized meanings or overly formal meanings unless clearly defined in the various embodiments of the present invention.

[0023] Example 1 The disclosed embodiment provides an underground UAV obstacle avoidance method, which is used to use point cloud data acquired by a laser radar to detect whether there are obstacles in a set area, so as to determine whether the UAV should brake to avoid the obstacle.

[0024] See also Figure 1 , an underground UAV obstacle avoidance method includes steps S101~S105, and each step is described in detail below.

[0025] Step S101, obtaining initial point cloud data corresponding to the underground environment, and preprocessing the initial point cloud data to obtain preprocessed point cloud data.

[0026] In this embodiment, a laser radar device is mounted on a drone to scan the underground environment in real time and obtain initial point cloud data of the surrounding environment; all initial point cloud data are further preprocessed to obtain high-quality preprocessed point cloud data to improve the computational efficiency of subsequent steps.

[0027] In a specific embodiment, step S101 includes: denoising the initial point cloud data based on a statistical filtering algorithm to obtain first point cloud data; filtering the first point cloud data based on a voxel filtering algorithm to obtain the preprocessed point cloud data.

[0028] In this embodiment, the initial point cloud data is denoised by a statistical filtering algorithm to obtain first point cloud data with noise points removed. At the same time, the denoised first point cloud data is filtered by a voxel filtering algorithm to obtain preprocessed point cloud data with redundant data removed. At this time, the number of point clouds in the data is significantly reduced, thereby providing high-quality input for subsequent mapping.

[0029] Step S102, based on a synchronous positioning and map building algorithm, construct a three-dimensional underground map according to the pre-processed point cloud data.

[0030] In this embodiment, based on the simultaneous localization and mapping (SLAM) algorithm, a three-dimensional underground map is constructed according to the pre-processed point cloud data to provide a spatial reference for the UAV.

[0031] Step S103, determining a non-ground point cloud according to the underground three-dimensional map, and extracting obstacle features from the non-ground point cloud.

[0032] In this embodiment, the underground three-dimensional map is used to divide the non-ground point cloud, and the obstacle features are detected and extracted from the non-ground point cloud. The obstacle features include information such as the position, size and shape of the obstacle.

[0033] Exemplarily, the parameters are initialized, specifically setting the size of the filter window to 3x3, setting the elevation threshold, and setting the number of iterations.

[0034] Furthermore, the point cloud data in the underground three-dimensional map are sorted by elevation, and a certain number of points are selected as initial ground seed points starting from the lowest point.

[0035] Furthermore, a filter window is used to slide on the point cloud data, and a morphological opening operation is performed on the points in each window. The elevation difference of the point cloud before and after the opening operation is compared. If the difference is less than the elevation threshold, the point is considered to be a ground point, and the newly identified ground point is added to the ground point set to update the ground point set. After each iteration, the new ground point set is used for the next round of iteration, that is, this step is repeated.

[0036] Furthermore, when the number of iterations reaches a preset number of iterations or no new ground points are identified in three consecutive iterations, the iterations are stopped.

[0037] Furthermore, the ground points are separated from the non-ground points, and the ground point cloud and the non-ground point cloud are output.

[0038] Furthermore, region growing is performed based on the intensity features of the point cloud to segment the non-ground point cloud into different regions.

[0039] Furthermore, the K-means clustering algorithm is used to divide the point clouds in different areas into multiple clusters, each cluster may represent an obstacle.

[0040] Furthermore, the clusters identified as obstacles are subjected to boundary detection to determine the exact outline of the obstacles.

[0041] Furthermore, points that do not belong to obstacles during the obstacle model fitting process, such as abnormal points or misclassified points, are removed to achieve the denoising function.

[0042] Furthermore, obstacle feature parameters, such as center position, size, and direction, are calculated based on the denoised obstacle point cloud data.

[0043] Step S104, determining the target obstacle avoidance area according to the drone positioning data and the remote controller command data.

[0044] In this embodiment, a designated target obstacle avoidance area is determined based on the drone positioning data and preset remote control command data. The target obstacle avoidance area is a key spatial range specially set to avoid collision with obstacles based on the current state of the drone and the operator's intention.

[0045] Step S105, performing obstacle avoidance for the drone according to the target obstacle avoidance area and the obstacle characteristics.

[0046] In this embodiment, by detecting whether there are obstacles in the designated target obstacle avoidance area, it is decided whether the UAV should brake, thereby realizing effective identification and avoidance of various obstacles by utilizing obstacle features obtained based on the point cloud data of the underground environment and the set obstacle avoidance area, thereby significantly improving the safety and autonomous navigation capability of the UAV in complex underground environments.

[0047] In a specific embodiment, step S104 includes: based on the size of the component data of the expected speed of the drone, determining the target obstacle avoidance area according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas.

[0048] In this embodiment, whether the expected movement speed of the drone in each direction is significant can be determined based on the size of the component data of the drone's expected speed, thereby determining the target obstacle avoidance area by combining the short-range horizontal obstacle avoidance area, the long-range horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas.

[0049] See also Figure 2 In a specific embodiment, the UAV positioning data includes current position data and current azimuth, and the remote control command data includes the UAV expected speed, the UAV horizontal expected obstacle avoidance distance, the UAV vertical expected obstacle avoidance distance, and the UAV maximum obstacle avoidance distance; the method also includes steps S201~S204, and each step is described in detail below.

[0050] Step S201, calculating the short-distance horizontal obstacle avoidance area according to the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone.

[0051] In this embodiment, based on the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data, and the expected speed of the drone, the horizontal obstacle avoidance area of ​​the drone at close range is determined through geometric and dynamic analysis to prevent collision with close-range obstacles.

[0052] See also Figure 3 In a specific embodiment, step S201 includes steps S2011 to S2014, and each step is described in detail below.

[0053] Step S2011, dynamically adjusting the safety distance according to the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone.

[0054] In this embodiment, the safety distance is dynamically adjusted based on the expected obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone. The drone body data includes the size data of the drone body. , the expected speed of the drone .

[0055] Specifically, dynamically adjust the safety distance , where is the expected obstacle avoidance distance of the UAV, is a fixed coefficient.

[0056] Step S2012, determining a preset fan-shaped body angle according to the tangent function and the expected speed of the drone.

[0057] In this embodiment, the preset fan-shaped body angle is determined according to the tangent function and the expected speed of the drone. The preset fan-shaped body can be understood as a fan-shaped body with the height of the drone as the height and the horizontal expected obstacle avoidance distance as the radius. The preset fan-shaped body angle represents the orientation angle of the fan-shaped body relative to the drone fuselage. Specifically, the preset fan-shaped body angle .

[0058] Step S2013, determining the vertical spatial range of the UAV according to the UAV fuselage data.

[0059] In this embodiment, the vertical spatial range of the drone is determined based on the drone body data. .

[0060] Step S2014, determining the short-range horizontal obstacle avoidance area according to the current position data, the dynamically adjusted safety distance, the preset fan-shaped body angle and the vertical spatial range of the UAV.

[0061] In this embodiment, the short-distance horizontal obstacle avoidance area is determined according to the current position data, the dynamically adjusted safety distance, the preset fan-shaped body angle and the vertical spatial range of the drone, wherein the current position data is . Specifically, the short-distance horizontal obstacle avoidance area .

[0062] Step S202, calculating the long-distance horizontal obstacle avoidance area according to the current position data, the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance.

[0063] In this embodiment, the horizontal obstacle avoidance area of ​​the drone within a long distance is determined based on the current position data, the current azimuth, the drone body data, the drone's expected horizontal obstacle avoidance distance, and the drone's maximum obstacle avoidance distance to avoid potential long-distance obstacles.

[0064] See also Figure 4 In a specific embodiment, step S202 includes steps S2021 to S2022, and each step is described in detail below.

[0065] Step S2021, based on the sine function and the cosine function, determine the obstacle avoidance range rectangular boundary and the rotation matrix according to the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance.

[0066] In this embodiment, based on the sine function and cosine function, the obstacle avoidance range rectangle boundary and the rotation matrix are determined according to the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance. Among them, the obstacle avoidance range rectangle is the obstacle avoidance area outside the short-distance obstacle avoidance range, and the purpose of the rotation matrix is ​​to rotate the obstacle avoidance area to the fuselage coordinate system.

[0067] Specifically, the obstacle avoidance range rectangular boundary is used express.

[0068] ∈

[0069] ∈

[0070]

[0071] In the formula, The current azimuth Counterclockwise rotation Angle, The current azimuth Rotate clockwise Angle, .

[0072] The rotation matrix is .

[0073] Step S2022, calculating the long-distance horizontal obstacle avoidance area according to the current position data, the obstacle avoidance range rectangular boundary and the rotation matrix.

[0074] In this embodiment, the long-distance horizontal obstacle avoidance area is calculated based on the current position data, the rectangular boundary of the obstacle avoidance range and the rotation matrix. , where is the current azimuth.

[0075] Step S203, calculating the vertical obstacle avoidance area according to the current position data and the drone body data.

[0076] In this embodiment, the obstacle avoidance area of ​​the drone in the vertical direction is determined based on the current position data and the drone body data to prevent collision with obstacles above or below.

[0077] See also Figure 5 In a specific embodiment, step S203 includes steps S2031 to S2032, and each step is described in detail below.

[0078] Step S2031, determining a preset cubic range according to the drone body data and the vertical expected obstacle avoidance distance of the drone.

[0079] In this embodiment, the preset cube range is determined according to the drone body data and the vertical expected obstacle avoidance distance of the drone. The preset cube is an upward or downward cube centered on the drone. Specifically, the preset cube range is , , , , where is the expected vertical obstacle avoidance distance of the UAV.

[0080] Step S2032: Calculate the vertical obstacle avoidance area according to the current position data and the preset cube range.

[0081] In this embodiment, the vertical obstacle avoidance area is calculated based on the current position data and the preset cube range. .

[0082] Step S204, calculating the obstacle avoidance areas in the horizontal and vertical directions according to the current position data, the current azimuth, the expected speed of the drone, and the expected horizontal obstacle avoidance distance of the drone.

[0083] In this embodiment, the current position, azimuth, expected speed and horizontal expected obstacle avoidance distance of the drone are combined, and the omnidirectional obstacle area that the drone may encounter during flight is calculated through three-dimensional spatial geometry and dynamic analysis. This area will take into account obstacles in both horizontal and vertical directions to ensure the safe flight of the drone.

[0084] See also Figure 6 In a specific embodiment, step S204 includes steps S2041 to S2042, and each step is described in detail below.

[0085] Step S2041, based on the sine function and the cosine function, the preset sphere local area range is determined according to the dynamically adjusted safety distance and the preset sector angle.

[0086] In this embodiment, based on the sine function and the cosine function, the preset sphere local area range is determined according to the dynamically adjusted safety distance and the preset sector angle. The preset sphere local area range represents 1 / 4 sphere. The preset sphere local area range is , where , Indicates the pitch range of the ball.

[0087] Step S2042, calculating the obstacle avoidance areas in the horizontal and vertical directions according to the current position data and the preset spherical local area range.

[0088] In this embodiment, the obstacle avoidance areas in the horizontal and vertical directions are calculated based on the current position data and the preset local area range of the sphere.

[0089] Horizontal and vertical obstacle avoidance areas It is expressed as:

[0090] See also Figure 7 In a specific embodiment, the desired speed of the drone includes an x ​​speed component, a y speed component and a z speed component. The component data size based on the desired speed of the drone is used to determine the target obstacle avoidance area according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area and the horizontal and vertical obstacle avoidance areas, including steps S701 to S705. Each step is described in detail below.

[0091] Step S701: If the absolute value of the x velocity component is greater than 0 and the absolute value of the z velocity component is within a preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area.

[0092] In this embodiment, if the absolute value of the x-velocity component (i.e., the velocity in the front-to-back direction) of the drone is greater than 0, and the absolute value of the z-velocity component (i.e., the velocity in the vertical direction) is within a preset range, the close-range horizontal obstacle avoidance area will be set. and long-distance horizontal obstacle avoidance area The sum of the values ​​is used as the target obstacle avoidance area. For a schematic diagram of the target obstacle avoidance area at this time, please refer to Figure 8 .

[0093] Because the drone is mainly moving in the front and back directions and is almost stationary in the vertical direction, the main obstacle avoidance consideration is in the horizontal direction. Among them, the preset range is close to 0.

[0094] Step S702: If the absolute value of the y velocity component is greater than 0 and the absolute value of the z velocity component is within the preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area.

[0095] In this embodiment, if the absolute value of the y velocity component (i.e., the velocity in the left and right directions) of the drone is greater than 0, and the absolute value of the z velocity component is within a preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is also used as the target obstacle avoidance area. Because the drone is mainly moving in the left and right directions at this time, the obstacle avoidance considerations are still mainly in the horizontal direction.

[0096] Step S703: If the absolute value of the x velocity component is within the preset range, the absolute value of the y velocity component is approximately 0, and the absolute value of the z velocity component is greater than 0, the vertical obstacle avoidance area is used as the target obstacle avoidance area.

[0097] In this embodiment, if the absolute values ​​of the x-velocity component and the y-velocity component of the drone are both approximately 0, and the absolute value of the z-velocity component is greater than 0, the vertical obstacle avoidance area is used as the target obstacle avoidance area. Because the drone mainly moves in the vertical direction (ascending or descending) and is almost stationary in the horizontal direction, the main obstacle avoidance consideration is in the vertical direction.

[0098] Step S704: If the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas.

[0099] In this embodiment, if the absolute value of the x-velocity component of the drone is greater than 0, and the absolute value of the z-velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as target obstacle avoidance areas. At this time, the drone moves in the front-back direction and the vertical direction at the same time, and the obstacle avoidance areas in the horizontal and vertical directions need to be considered at the same time.

[0100] Step S705: If the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas.

[0101] In this embodiment, if the absolute value of the y velocity component of the drone is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are also used as target obstacle avoidance areas. At this time, the drone moves in the left and right directions and in the vertical direction at the same time, and the obstacle avoidance considerations need to cover both the horizontal and vertical directions.

[0102] The underground drone obstacle avoidance method proposed in this embodiment obtains the initial point cloud data corresponding to the underground environment, and pre-processes the initial point cloud data to obtain pre-processed point cloud data; constructs an underground three-dimensional map based on the pre-processed point cloud data based on the synchronous positioning and map construction algorithm; determines the non-ground point cloud based on the underground three-dimensional map, and extracts obstacle features from the non-ground point cloud; determines the target obstacle avoidance area based on the drone positioning data and the remote control command data; and performs drone obstacle avoidance based on the target obstacle avoidance area and the obstacle features. In this way, the point cloud data in the underground environment is obtained by the laser radar, and the obstacle features are extracted from it; at the same time, the target obstacle avoidance area is determined based on the drone positioning data and the remote control command data, so as to combine the obstacle features and the target obstacle avoidance area to avoid obstacles, and then even in the dark underground environment, it can effectively identify and avoid various obstacles, ensuring the safety and smoothness of the drone's movement.

[0103] Example 2 In addition, the present disclosure provides an underground drone obstacle avoidance device 900, see Fig. 9 , the device comprises: The acquisition module 901 is used to acquire initial point cloud data corresponding to the downhole environment, and pre-process the initial point cloud data to obtain pre-processed point cloud data; A construction module 902 is used to construct a three-dimensional underground map according to the pre-processed point cloud data based on a synchronous positioning and map construction algorithm; An extraction module 903, configured to determine a non-ground point cloud according to the underground three-dimensional map, and extract obstacle features from the non-ground point cloud; A determination module 904 is used to determine a target obstacle avoidance area according to the drone positioning data and the remote controller command data; The obstacle avoidance module 905 is used to perform obstacle avoidance for the drone according to the target obstacle avoidance area and the obstacle characteristics.

[0104] Optionally, the drone positioning data includes current position data and current azimuth, and the remote control command data includes the drone expected speed, the drone horizontal expected obstacle avoidance distance, the drone vertical expected obstacle avoidance distance, and the drone maximum obstacle avoidance distance; the determination module 904 is used to determine the target obstacle avoidance area based on the size of the component data of the drone expected speed according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas; wherein the short-distance horizontal obstacle avoidance area is calculated according to the current position data, the drone horizontal expected obstacle avoidance distance, the drone fuselage data, and the drone expected speed; the long-distance horizontal obstacle avoidance area is calculated according to the current position data, the current azimuth, the drone fuselage data, the drone horizontal expected obstacle avoidance distance, and the drone maximum obstacle avoidance distance; the vertical obstacle avoidance area is calculated according to the current position data, the drone fuselage data, and the drone vertical expected obstacle avoidance distance; the horizontal and vertical obstacle avoidance areas are calculated according to the current position data, the current azimuth, the drone expected speed, and the drone horizontal expected obstacle avoidance distance.

[0105] Optionally, the expected speed of the drone includes an x-speed component, a y-speed component, and a z-speed component. The determination module 904 is further used to, if the absolute value of the x-speed component is greater than 0 and the absolute value of the z-speed component is within a preset range, use the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area as the target obstacle avoidance area; if the absolute value of the y-speed component is greater than 0 and the absolute value of the z-speed component is within the preset range, use the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area as the target obstacle avoidance area. area; if the absolute value of the x velocity component is within the preset range, the absolute value of the y velocity component is within the preset range, and the absolute value of the z velocity component is greater than 0, the vertical obstacle avoidance area is used as the target obstacle avoidance area; if the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas; if the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas.

[0106] Optionally, a determination module 904 is used to calculate a dynamically adjusted safety distance based on the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone; determine a preset fan-shaped angle based on the tangent function and the expected speed of the drone; determine the vertical spatial range of the drone based on the drone body data; and determine the close-range horizontal obstacle avoidance area based on the current position data, the dynamically adjusted safety distance, the preset fan-shaped angle and the vertical spatial range of the drone.

[0107] Optionally, a determination module 904 is used to determine the rectangular boundary and rotation matrix of the obstacle avoidance range based on the sine function and the cosine function according to the current azimuth angle, the drone body data, the expected horizontal obstacle avoidance distance of the drone and the maximum obstacle avoidance distance of the drone; and calculate the long-distance horizontal obstacle avoidance area according to the current position data, the rectangular boundary of the obstacle avoidance range and the rotation matrix.

[0108] Optionally, the determination module 904 is used to determine a preset cube range according to the drone body data and the vertical expected obstacle avoidance distance of the drone; and calculate the vertical obstacle avoidance area according to the current position data and the preset cube range.

[0109] Optionally, a determination module 904 is used to determine the preset sphere local area range based on the sine function and the cosine function according to the dynamically adjusted safety distance and the preset fan-shaped angle; and calculate the horizontal and vertical obstacle avoidance areas according to the current position data and the preset sphere local area range.

[0110] Optionally, the acquisition module 901 is used to denoise the initial point cloud data based on a statistical filtering algorithm to obtain first point cloud data; and filter the first point cloud data based on a voxel filtering algorithm to obtain the preprocessed point cloud data.

[0111] The device provided in the embodiment of the present disclosure can execute the steps of the underground drone obstacle avoidance method provided in Example 1, which will not be described again to avoid repetition.

[0112] The underground drone obstacle avoidance device proposed in this embodiment obtains the initial point cloud data corresponding to the underground environment, and pre-processes the initial point cloud data to obtain pre-processed point cloud data; constructs an underground three-dimensional map based on the pre-processed point cloud data based on the synchronous positioning and map construction algorithm; determines the non-ground point cloud based on the underground three-dimensional map, and extracts obstacle features from the non-ground point cloud; determines the target obstacle avoidance area based on the drone positioning data and the remote control command data; and performs drone obstacle avoidance based on the target obstacle avoidance area and the obstacle features. In this way, the point cloud data in the underground environment is obtained by the laser radar, and the obstacle features are extracted therefrom; at the same time, the target obstacle avoidance area is determined based on the drone positioning data and the remote control command data, so as to combine the obstacle features and the target obstacle avoidance area to avoid obstacles, and then, even in a dark underground environment, various obstacles can be effectively identified and avoided, ensuring the safety and smoothness of the drone's movement.

[0113] Example 3 In addition, an embodiment of the present disclosure provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the underground drone obstacle avoidance method described in Example 1 is implemented.

[0114] The device provided in the embodiment of the present disclosure can execute the steps of the underground drone obstacle avoidance method provided in Example 1, which will not be described again to avoid repetition.

[0115] Example 4 The disclosed embodiment provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the underground drone obstacle avoidance method described in Embodiment 1.

[0116] In this embodiment, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0117] The computer-readable storage medium provided in this embodiment can implement the underground drone obstacle avoidance method provided in Example 1. To avoid repetition, it will not be repeated here.

[0118] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limiting, and thus other examples of the exemplary embodiments may have different values.

[0119] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0120] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. An underground drone obstacle avoidance method, characterized in that: include: Acquiring initial point cloud data corresponding to the underground environment, and preprocessing the initial point cloud data to obtain preprocessed point cloud data; Based on the synchronous positioning and mapping algorithm, construct a three-dimensional map of the well according to the pre-processed point cloud data; Determine a non-ground point cloud according to the underground three-dimensional map, and extract obstacle features from the non-ground point cloud; Determine the target obstacle avoidance area based on the drone positioning data and remote controller command data; The drone avoids obstacles according to the target obstacle avoidance area and the obstacle characteristics.

2. The underground drone obstacle avoidance method according to claim 1, characterized in that: The drone positioning data includes current position data and current azimuth, and the remote control command data includes the drone expected speed, the drone horizontal expected obstacle avoidance distance, the drone vertical expected obstacle avoidance distance, and the drone maximum obstacle avoidance distance; Determining the target obstacle avoidance area according to the drone positioning data and the remote controller command data includes: Based on the size of the component data of the expected speed of the drone, the target obstacle avoidance area is determined according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas; The short-distance horizontal obstacle avoidance area is calculated according to the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone; Calculate the long-distance horizontal obstacle avoidance area according to the current position data, the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance; Calculate the vertical obstacle avoidance area according to the current position data, the drone body data and the vertical expected obstacle avoidance distance of the drone; The obstacle avoidance areas in the horizontal and vertical directions are calculated according to the current position data, the current azimuth angle, the expected speed of the drone and the expected horizontal obstacle avoidance distance of the drone.

3. The underground drone obstacle avoidance method according to claim 2, characterized in that: The desired speed of the drone includes an x ​​speed component, a y speed component, and a z speed component. The target obstacle avoidance area is determined based on the size of the component data of the desired speed of the drone according to the short-distance horizontal obstacle avoidance area, the long-distance horizontal obstacle avoidance area, the vertical obstacle avoidance area, and the horizontal and vertical obstacle avoidance areas, including: If the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is within a preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is within the preset range, the sum of the short-distance horizontal obstacle avoidance area and the long-distance horizontal obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the x velocity component is within the preset range, the absolute value of the y velocity component is within the preset range, and the absolute value of the z velocity component is greater than 0, the vertical obstacle avoidance area is used as the target obstacle avoidance area; If the absolute value of the x velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas; If the absolute value of the y velocity component is greater than 0, and the absolute value of the z velocity component is greater than 0, the horizontal and vertical obstacle avoidance areas are used as the target obstacle avoidance areas.

4. The underground drone obstacle avoidance method according to claim 2, characterized in that: The calculating the short-distance horizontal obstacle avoidance area according to the current position data, the expected horizontal obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone includes: Calculate and dynamically adjust the safety distance according to the expected obstacle avoidance distance of the drone, the drone body data and the expected speed of the drone; Determine a preset sector angle according to the tangent function and the desired speed of the drone; Determine the vertical spatial range of the drone according to the drone fuselage data; The short-range horizontal obstacle avoidance area is determined according to the current position data, the dynamically adjusted safety distance, the preset fan-shaped body angle and the vertical spatial range of the UAV.

5. The underground drone obstacle avoidance method according to claim 2, characterized in that: The calculating the long-distance horizontal obstacle avoidance area according to the current position data, the current azimuth, the drone body data, the drone horizontal expected obstacle avoidance distance and the drone maximum obstacle avoidance distance includes: Based on the sine function and the cosine function, determining the obstacle avoidance range rectangular boundary and the rotation matrix according to the current azimuth angle, the drone body data, the horizontal expected obstacle avoidance distance of the drone and the maximum obstacle avoidance distance of the drone; The long-distance horizontal obstacle avoidance area is calculated according to the current position data, the obstacle avoidance range rectangular boundary and the rotation matrix.

6. The underground drone obstacle avoidance method according to claim 2, characterized in that: The calculating the vertical obstacle avoidance area according to the current position data, the drone body data and the vertical expected obstacle avoidance distance of the drone includes: Determine a preset cube range according to the drone fuselage data and the vertical expected obstacle avoidance distance of the drone; The vertical obstacle avoidance area is calculated according to the current position data and the preset cube range.

7. The underground drone obstacle avoidance method according to claim 4, characterized in that: The calculating the obstacle avoidance areas in the horizontal and vertical directions according to the current position data, the current azimuth, the expected speed of the drone, and the expected horizontal obstacle avoidance distance of the drone includes: Based on the sine function and the cosine function, the preset sphere local area range is determined according to the dynamically adjusted safety distance and the preset sector angle; The obstacle avoidance areas in the horizontal and vertical directions are calculated according to the current position data and the preset local area range of the sphere.

8. The underground drone obstacle avoidance method according to claim 1, characterized in that: The preprocessing of the initial point cloud data to obtain preprocessed point cloud data includes: De-noising the initial point cloud data based on a statistical filtering algorithm to obtain first point cloud data; The first point cloud data is filtered based on a voxel filtering algorithm to obtain the preprocessed point cloud data.

9. A computer device, characterized in that: It comprises a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the underground UAV obstacle avoidance method as described in any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium, characterized in that: It stores a computer program, which, when executed by a processor, implements the underground UAV obstacle avoidance method as described in any one of claims 1 to 8.

Images