CN-121363952-B - Unmanned aerial vehicle fusion map building and positioning method and device, computer equipment and storage medium

CN121363952BCN 121363952 BCN121363952 BCN 121363952BCN-121363952-B

Abstract

The invention relates to a fusion map building and positioning method and device of an unmanned aerial vehicle, computer equipment and a storage medium. The unmanned aerial vehicle is provided with a solid-state laser radar and a rotatable support, the rotatable support drives the solid-state laser radar to rotate, the method comprises the steps of obtaining an original laser point cloud data frame according to the solid-state laser radar, obtaining IMU data according to an IMU of a body of the unmanned aerial vehicle, preprocessing the original laser point cloud data frame and the IMU data to obtain preprocessed data, obtaining a pose of the unmanned aerial vehicle according to the preprocessed data, and obtaining a double-layer map according to the pose of the unmanned aerial vehicle and the preprocessed data, wherein the double-layer map comprises a sparse feature map and a visual dense body map. The invention can improve the precision of the unmanned aerial vehicle and reduce the quality of the unmanned aerial vehicle.

Inventors

WANG JUNYAO
SONG CHUAN
LI GUO
GAO MENGJIAO
CHENG LIANJUN
LEI LIN

Assignees

珠海华发人居生活研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20251222

Claims (12)

1. The utility model provides a unmanned aerial vehicle fuses and builds drawing and positioning method which characterized in that, be provided with solid-state laser radar and rotatable support on the unmanned aerial vehicle, rotatable support drives solid-state laser radar rotates, the method includes: Acquiring an original laser point cloud data frame according to the solid-state laser radar; Acquiring IMU data according to an engine body IMU of the unmanned aerial vehicle; preprocessing the original laser point cloud data frame and the IMU data to obtain preprocessed data; acquiring the pose of the unmanned aerial vehicle according to the preprocessed data; Acquiring a double-layer map according to the pose of the unmanned aerial vehicle and the preprocessed data; wherein the dual-layer map comprises a sparse feature map and a visualized dense voxel map; the obtaining the original laser point cloud data frame according to the solid-state laser radar comprises the following steps: dividing the visual field corresponding to one circle of rotation of the solid-state laser radar into a plurality of angular windows according to the phase of the rotatable bracket, taking the boundary between two adjacent angular windows as an angular window boundary, The solid-state laser radar rotates under the drive of the rotatable bracket and acquires an original point cloud, Dividing the original point cloud into a plurality of quasi-helical subframes according to the angular window boundary, Wherein the plurality of quasi-helical subframes constitute the original laser point cloud data frame; the width constraint condition of the corner window is as follows: The phase span of the angular window is as follows: Wherein, the For the width of the angular window, For the maximum linear displacement to be the maximum, For the average linear velocity of the rotating electrical machine, For the maximum angular displacement to be the maximum, For the average angular velocity of the rotating electrical machine, For the electrical frequency of the laser radar, For the minimum number of single quasi-helical subframes, For the phase span the phase of the optical signal is shifted, Is the first The starting phase of each quasi-helical subframe, Is the first The starting phase of each quasi-helical subframe.
2. The method of claim 1, wherein the angular window width constraint is: The linear displacement of each of the quasi-helical subframes does not exceed a maximum linear displacement, and the angular displacement of each of the quasi-helical subframes does not exceed a maximum angular displacement.
3. The method of claim 1, wherein dividing the field of view corresponding to one rotation of the solid state lidar into a plurality of angular windows according to the phase of the rotatable support comprises: according to the shaking degree of the unmanned aerial vehicle during flight, the width of the angular window is adaptively adjusted; Wherein the greater the degree of jitter during flight, the smaller the width of the angular window.
4. The method of claim 1, wherein the original laser point cloud data frame and the IMU data have time stamps, The preprocessing the original laser point cloud data frame and the IMU data to obtain preprocessed data comprises the following steps: Taking each angular window boundary as the ending phase of the current quasi-spiral subframe and the starting phase of the next quasi-spiral subframe, and taking the starting phase of the current quasi-spiral subframe as a reference moment; according to the time stamp, the original point cloud in each quasi-spiral subframe is callback to the corresponding reference moment according to the IMU data, and a callback data frame is obtained; According to the IMU rotation rate, compensating the time deviation of the callback data frame to remove distortion in the point cloud data and obtain a de-distorted data frame; Filtering the de-distorted data frame according to voxel filtering, statistical group point removing filtering and intensity threshold filtering to obtain preprocessed data; Wherein the callback data frame is aligned with the phase of the IMU data.
5. The method of claim 4, wherein compensating the time offset of the callback data frame according to the IMU rotation rate to remove distortion in the point cloud data and obtain a de-distorted data frame comprises: Acquiring the angular speed of an encoder according to the rotation speed of the IMU; Acquiring the relative pose of the point cloud of the callback data frame at the reference moment of the quasi-spiral subframe according to the pre-integration of the IMU data and the angular speed of the encoder; and transforming the point clouds of all the callback data frames to a radar coordinate system or an organism coordinate system through coordinate transformation to obtain corrected point clouds with consistent space time as the de-distorted data frames.
6. The method of claim 5, wherein transforming the point clouds of all the callback data frames to a radar coordinate system or a body coordinate system by coordinate transformation comprises: obtaining an external parameter matrix between the solid-state laser radar and the IMU of the machine body; Acquiring the current angle of the rotatable bracket; And according to the current angle of the rotatable bracket and the current external parameter matrix, mapping the point cloud of the callback data frame from the radar coordinate system to the machine body coordinate system or from the machine body coordinate system to the radar coordinate system.
7. The method of claim 1, wherein the acquiring the pose of the unmanned aerial vehicle from the preprocessed data comprises: acquiring an inter-frame motion initial value according to the pre-integration of the IMU data, and acquiring state prediction according to the inter-frame motion initial value; according to the state prediction, the current quasi-spiral subframe and the adjacent quasi-spiral subframe are quickly registered to obtain a coarse pose; According to the rough pose, a plurality of continuous quasi-spiral subframes are aggregated to form a local sub-map; and carrying out fine registration on the local sub map and the historical sub map so as to obtain an accurate pose.
8. The method of claim 7, wherein the obtaining a dual-layer map from the pose of the drone and the preprocessed data comprises: According to the state prediction, the current quasi-spiral subframe and the adjacent quasi-spiral subframe are quickly registered, and the sparse feature map is obtained, wherein the sparse feature map comprises line features, surface features, feature covariance, phase coverage histograms of features and observation times; According to the rough pose, a plurality of continuous quasi-spiral subframes are aggregated to form a local sub-map; And carrying out fine registration on the local sub-map and the historical subgraph to obtain the visual dense voxel map.
9. The method of claim 8, wherein the acquiring a dual-layer map from the pose of the drone and the preprocessed data, further comprises: tightly coupling and fusing the coarse pose and IMU prediction in iterative error state Kalman filtering to obtain an optimal state; and according to the optimal state, projecting the current key frame in the iterative error state Kalman filtering to a global coordinate system, and fusing the current key frame into a previous double-layer map to obtain the current double-layer map.
10. A fusion mapping and positioning device for an unmanned aerial vehicle, characterized in that a solid-state laser radar and a rotatable bracket are arranged on the unmanned aerial vehicle, the rotatable bracket drives the solid-state laser radar to rotate, and the device applies the method as claimed in any one of claims 1 to 9, and the device comprises: The point cloud acquisition unit is used for acquiring an original laser point cloud data frame according to the solid-state laser radar; the IMU data acquisition unit is used for acquiring IMU data according to the IMU of the unmanned aerial vehicle body; The preprocessing unit is used for preprocessing the original laser point cloud data frame and the IMU data to obtain preprocessed data; the pose acquisition unit is used for acquiring the pose of the unmanned aerial vehicle according to the preprocessed data; The map acquisition unit is used for acquiring a double-layer map according to the pose of the unmanned aerial vehicle and the preprocessed data; wherein the dual-layer map comprises a sparse feature map and a visualized dense voxel map.
11. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any one of claims 1 to 9 when executing the computer program.
12. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the method of any one of claims 1 to 9.

Description

Unmanned aerial vehicle fusion map building and positioning method and device, computer equipment and storage medium Technical Field The invention relates to the field of unmanned aerial vehicles, in particular to a method, a device, computer equipment and a storage medium for fusion mapping and positioning of unmanned aerial vehicles. Background Unmanned aerial vehicle's flight, location, map construction etc. rely on the cooperation of laser radar and other equipment. Lidars are generally classified into rotary lidars and solid-state lidars. Rotary office radars are typically 32-line or 64-line lasers that achieve 360-degree scanning of the laser light by rotation of the laser inside the laser radar. Because of the volume of the laser transmitter in the laser radar, the volume of the optical equipment matched with the laser radar and the circuit volume, the rotary laser radar has great difficulty in realizing more lines of beams, and the resolution of the rotary laser radar is restricted. Solid-state lidars generally realize laser scanning by vibrating mirrors at high speed, so that extremely high resolution is possible, but the detection field of the solid-state lidar is limited, usually about 120 degrees, and comprehensive detection is difficult to realize. To address the limitation of the field of view of solid state lidars, multiple solid state lidars may be provided to achieve a 360 field of view. However, the solid-state laser radar has a large volume and a heavy mass, and the cost is far higher than that of the rotary laser radar, so that it is difficult to configure a plurality of solid-state laser radars on the unmanned aerial vehicle. Therefore, the unmanned aerial vehicle in the prior art is difficult to realize high-precision fusion map building and positioning through the solid-state laser radar. Disclosure of Invention In order to solve the technical problems or at least partially solve the technical problems, the invention provides a fusion map and positioning method, a device, computer equipment and a storage medium of an unmanned aerial vehicle. In a first aspect, the present invention provides a fusion mapping and positioning method for an unmanned aerial vehicle, where the unmanned aerial vehicle is provided with a solid-state laser radar and a rotatable bracket, and the rotatable bracket drives the solid-state laser radar to rotate, and the method includes: Acquiring an original laser point cloud data frame according to the solid-state laser radar; Acquiring IMU data according to an engine body IMU of the unmanned aerial vehicle; preprocessing the original laser point cloud data frame and the IMU data to obtain preprocessed data; acquiring the pose of the unmanned aerial vehicle according to the preprocessed data; Acquiring a double-layer map according to the pose of the unmanned aerial vehicle and the preprocessed data; wherein the dual-layer map comprises a sparse feature map and a visualized dense voxel map. Optionally, the acquiring the original laser point cloud data frame according to the solid-state laser radar includes: dividing a visual field corresponding to one circle of rotation of the solid-state laser radar into a plurality of angular windows according to the phase of the rotatable bracket, wherein the boundary between two adjacent angular windows is used as an angular window boundary; The solid-state laser radar rotates under the drive of the rotatable bracket and acquires an original point cloud; dividing the original point cloud into a plurality of quasi-spiral subframes according to the angular window boundary; Wherein the plurality of quasi-helical subframes constitute the original laser point cloud data frame. Optionally, the width constraint condition of the angular window is: The linear displacement of each of the quasi-helical subframes does not exceed a maximum linear displacement, and the angular displacement of each of the quasi-helical subframes does not exceed a maximum angular displacement. Optionally, the dividing the view corresponding to one rotation of the solid-state laser radar into a plurality of angular windows according to the phase of the rotatable bracket includes: according to the shaking degree of the unmanned aerial vehicle during flight, the width of the angular window is adaptively adjusted; Wherein the greater the degree of jitter during flight, the smaller the width of the angular window. Optionally, the width constraint condition of the angular window is: The phase span of the angular window is as follows: Wherein, the For the width of the angular window,For the maximum linear displacement to be the maximum,For the average linear velocity of the rotating electrical machine,For the maximum angular displacement to be the maximum,For the average angular velocity of the rotating electrical machine,For the electrical frequency of the laser radar,For the minimum number of single quasi-helical subframes,For the phase span the phase of the optic