A real-time POS boom compensation method for an airborne LiDAR system

Abstract

This invention discloses a real-time POS lever arm compensation method for an airborne LiDAR system, belonging to the field of UAV technology; it includes the following steps: S1, hardware deployment and nominal model establishment: two antennas are installed on the UAV, a virtual antenna is generated by calculation, and an initial geometric model is established; S2, open-loop virtual antenna calculation: the nominal coordinates of the virtual antenna are calculated using the original coordinates of the two antennas; S3, closed-loop data correction: the virtual antenna coordinates are output through EKF data fusion and state estimation for subsequent point cloud calculation; this invention reduces the effective lever arm by constructing a virtual antenna using geometry, and further estimates and compensates for residual lever arm errors caused by factors such as inaccurate model, minor installation changes, or flight deformation in real time by introducing a feedback correction mechanism, thereby achieving self-optimization and continuous maintenance of system accuracy.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a real-time POS lever compensation method for an airborne LiDAR system. Background Technology

[0002] Airborne lidar 3D imaging systems generate 3D point clouds by fusing 2D point cloud data acquired by lidar with high-precision position and attitude information provided by a positioning and orientation system to calculate the 3D spatial coordinates of each lidar point. Currently, POS systems, through differential base station technology and integration with a high-precision inertial navigation system, have achieved pose accuracy that basically meets the imaging requirements of conventional 3D point clouds.

[0003] In airborne lidar systems, due to installation layout limitations, the POS satellite antenna and inertial navigation unit (IMU) often cannot be directly connected via a rigid structure. They are typically mounted separately at different locations on the UAV, resulting in a significant physical linkage between them. When the UAV's attitude changes drastically, relative attitude movement occurs between the airframe and the lidar system, causing the actual linkage between the antenna and IMU to change dynamically. This change is random and introduces additional position and attitude errors, leading to a decrease in POS output accuracy and consequently affecting the final point cloud imaging quality and stitching accuracy.

[0004] In existing technologies, this problem is typically addressed by connecting the satellite antenna and radar with a fixed rigid structure to reduce relative motion. However, this method has significant drawbacks: the rigid connection structure may interfere with the normal operation of the UAV's obstacle avoidance radar, increase the system's wind resistance, affect flight stability, and introduce collision risks, posing safety hazards.

[0005] This invention proposes a compensation and correction method that does not affect the flight safety of UAVs and can effectively suppress the impact of stick arm changes on POS accuracy. Summary of the Invention

[0006] The purpose of this invention is to solve the problem in the prior art where changes in the lever arm between the satellite antenna and the IMU lead to a decrease in POS accuracy, which in turn affects the quality of the lidar point cloud. Therefore, this invention proposes a real-time POS lever arm compensation method for an airborne LiDAR system.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A real-time POS lever compensation method for an airborne LiDAR system includes the following steps:

[0009] S1. Hardware Deployment and Nominal Model Establishment:

[0010] Install two antennas on the drone and calculate to generate a virtual antenna. Establish an initial geometric model;

[0011] S2, Open-loop virtual antenna calculation:

[0012] Calculate the nominal coordinates of the virtual antenna using the original coordinates of the two antennas;

[0013] S3, Closed-loop data correction:

[0014] By fusing EKF data and estimating its state, the coordinates of the virtual antenna are output for subsequent point cloud calculations.

[0015] In some embodiments, in step S1:

[0016] Two satellite antennas , Installed on the rotor arm of a drone;

[0017] Establish the body coordinate system P with the laser radar mount installation center as the origin;

[0018] Precise measurement to obtain antenna , Nominal coordinates in the P-frame , ;

[0019] Calculate virtual antenna nominal coordinates and the nominal distance to the two actual antennas. , .

[0020] In some embodiments, in step S2:

[0021] At each navigation epoch, acquire the antenna , Original coordinates in navigation frame n , ;

[0022] The nominal output coordinates of the virtual antenna are calculated using the inverse distance weighting method. :

[0023] ;

[0024] ;

[0025] ;

[0026] in, , Antenna , The weight.

[0027] In some embodiments, in step S3: the lever arm deviation state quantity is expanded in the Kalman filter of the POS. ;use Based on the current estimated attitude, calculate and output the corrected virtual antenna coordinates for subsequent point cloud calculations.

[0028] In some embodiments, step S3 includes:

[0029] Modeling the actual and nominal positions of virtual antennas tiny offset between ;

[0030] The attitude matrix estimated based on the current filter and contain The nominal geometric model predicts the baseline vector as ;

[0031] virtual antenna nominal coordinates The virtual antenna predicted coordinates are compared with the IMU position predicted by the filter and the arm vector after the current attitude and arm state transformation. The difference also constitutes the observation.

[0032] Using the estimated Based on the current estimated attitude, calculate the projection of the lever arm deviation onto the navigation frame. The corrected virtual antenna coordinates are output as follows: This is used for subsequent point cloud calculations.

[0033] In some embodiments, during point cloud solving:

[0034] Based on the virtual antenna coordinates, coordinate correction is performed, and a high-precision POS is output for laser point cloud settlement.

[0035] In some embodiments, the method further includes: S4, parameter adaptation;

[0036] Will Feedback is sent to the geometric model to update the effective body coordinates of the virtual antenna and dynamically optimize the model itself.

[0037] In some embodiments, step S4 further includes: updating the weighting coefficients;

[0038] Weighting coefficient Slowly adaptive adjustments are made based on residual information.

[0039] Compared with the prior art, the present invention provides a real-time POS lever compensation method for an airborne LiDAR system, which has the following beneficial effects.

[0040] 1. This invention reduces the effective boom arm by constructing a virtual antenna using geometry, and further estimates and compensates for residual boom arm errors caused by factors such as inaccurate model, minor installation changes, or flight deformation in real time by introducing a feedback correction mechanism, thereby achieving self-optimization and continuous maintenance of system accuracy.

[0041] Other advantages, objectives and features of the invention will be set forth in part in the description which follows; and in part will be apparent to those skilled in the art upon examination of the following description; or may be learned from practice of the invention. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of a drone's virtual antenna.

[0044] Figure 2 This is a schematic diagram of the virtual antenna of a drone (front view).

[0045] Figure 3 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0047] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0048] Currently, the satellite antenna of airborne lidar measurement systems is typically mounted on top of the aircraft or at the outer end of a rotating arm, while the radar is mounted on the bottom of the aircraft via a shock-absorbing pylon, with an integrated inertial navigation system (INS). Due to the requirements of the installation location, the arm length between the INS and the antenna is generally large (often greater than 50 cm). When the aircraft's attitude changes rapidly, the antenna and INS will rotate relative to each other, causing changes in the arm length; moreover, the larger the arm length, the more significant the relative displacement between them. This change is random and leads to a decrease in the pose accuracy of the position measurement system (POS).

[0049] If the antenna can be installed at the center of rotation of the radar (usually at the connection between the mount and the radar), the relative change in the mast arm is negligible; however, in reality, if installed this way, the antenna will be obstructed.

[0050] Reference Figure 1-3 A real-time POS boom compensation method for an airborne LiDAR system, involving dynamic compensation and closed-loop feedback correction of the satellite antenna boom in an UAV airborne LiDAR system; including the following steps:

[0051] S1. Hardware Deployment and Nominal Model Establishment:

[0052] Install two antennas on the drone and calculate to generate a virtual antenna. Establish an initial geometric model;

[0053] S2, Open-loop virtual antenna calculation:

[0054] Calculate the nominal coordinates of the virtual antenna using the original coordinates of the two antennas;

[0055] S3, Closed-loop data correction:

[0056] By fusing EKF data and estimating its state, the coordinates of the virtual antenna are output for subsequent point cloud calculations.

[0057] S4, parameter adaptive;

[0058] Update the nominal geometric model and weighting coefficients.

[0059] This invention generates a virtual antenna by mounting two antennas on diagonally opposite rotor arms of a drone and employing closed-loop feedback correction; this greatly improves the accuracy of POS without affecting the drone.

[0060] (1) Hardware deployment and nominal model establishment.

[0061] like Figure 1 and Figure 2 As shown; two satellite antennas with identical performance (antenna 1 and antenna 2) are mounted on a diagonally symmetrical rotor arm of the UAV; a coordinate system P is established with the laser radar mount mounting center (approximately the rotation center O) as the origin; precise measurements are taken to obtain the antenna 1 ( Antenna 2 Nominal coordinates in the P system , And based on this, calculate the virtual antenna. Nominal coordinates in the P-frame And to two real antennas , nominal distance , .

[0062] This step establishes the initial geometric model of the system.

[0063] (2) Open-loop virtual antenna solution.

[0064] At each navigation epoch, acquire the antenna , Original coordinates in navigation frame n , ;

[0065] The nominal output coordinates of the virtual antenna are calculated using the inverse distance weighting method. :

[0066]

[0067] in, , Antenna , The weights;

[0068] ;

[0069] .

[0070] (3) Closed-loop feedback correction.

[0071] By fusing EKF data and estimating its state, the coordinates of the virtual antenna are output for subsequent point cloud calculations.

[0072] In the Kalman filter of POS, the state variables In addition to standard states such as position, velocity, attitude, and IMU zero bias, a lever deviation state quantity is added. This deviation can be modeled as the difference between the actual and nominal positions of the virtual antenna in the body coordinate system P. The tiny offset between them, i.e. .

[0073] By using the double-difference carrier phase observations measured by two real antennas, the antenna phase can be calculated with high precision. , Projection of baseline vector in navigation frame .

[0074] Meanwhile, based on the attitude matrix estimated by the current filter and nominal geometric model (including The baseline vector can be predicted as follows: The difference between the two constitutes a strong constraint on the observation of attitude and lever arm deviation.

[0075] The virtual antenna nominal coordinates calculated in the open loop Add the IMU position predicted by the filter to the current attitude and lever state (including) The virtual antenna predicted coordinates obtained from the transformed arm vector are compared, and the difference also constitutes the observation.

[0076] Using the estimated Based on the current estimated attitude, calculate the projection of the lever arm deviation onto the navigation frame. The final, corrected virtual antenna coordinates are output as follows: This is used for subsequent point cloud calculations.

[0077] (iv) Parameter adaptation.

[0078] The estimated Feedback is sent to the geometric model to update the effective body coordinates (nominal geometric model) of the virtual antenna; that is, in the next calculation, the geometric model will be used. It replaces the original nominal value and dynamically optimizes the model itself.

[0079] This also includes: updating the weighting coefficients;

[0080] Weighting coefficient It can also be slowly and adaptively adjusted based on residual information.

[0081] This invention proposes a real-time boom compensation method based on dual-antenna virtual synthesis and closed-loop feedback correction. This method not only reduces the effective boom length by geometrically constructing a virtual antenna, but also estimates and compensates for residual boom length errors caused by factors such as inaccurate model, minor installation changes, or flight deformation in real time by introducing a feedback correction mechanism, thereby achieving self-optimization and continuous maintenance of system accuracy.

[0082] Experiments have shown that after using this invention: the absolute value of the boom arm is reduced from 50cm to 10cm; when the UAV and radar undergo relative attitude movement (assuming an amplitude of 5°), the change in boom arm is reduced from 4.36cm to 0.87cm; the heading accuracy of real-time POS is improved from 0.15° to 0.05°; the accuracy of roll and pitch angles is improved from 0.04° to 0.01°; and the thickness and overlap of the three-dimensional point cloud are significantly improved.

[0083] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

[0084] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0085] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A real-time POS lever compensation method for an airborne LiDAR system, characterized in that, Includes the following steps: S1. Hardware Deployment and Nominal Model Establishment: Install two antennas on the drone and calculate to generate a virtual antenna. Establish an initial geometric model; S2, Open-loop virtual antenna calculation: Calculate the nominal coordinates of the virtual antenna using the original coordinates of the two antennas; S3, Closed-loop data correction: By fusing EKF data and estimating its state, the coordinates of the virtual antenna are output for subsequent point cloud calculations.

2. The real-time POS lever compensation method for an airborne LiDAR system according to claim 1, characterized in that, In step S1: Two satellite antennas , Installed on the rotor arm of a drone; Establish the body coordinate system P with the laser radar mount installation center as the origin; Precise measurement to obtain antenna , nominal coordinates in the P-frame , ; Calculate virtual antenna nominal coordinates and the nominal distance to the two actual antennas. , .

3. The real-time POS lever compensation method for an airborne LiDAR system according to claim 2, characterized in that, In step S2: At each navigation epoch, acquire the antenna , Original coordinates in navigation frame n , ; The nominal output coordinates of the virtual antenna are calculated using the inverse distance weighting method. : in, , Antenna , The weight.

4. The real-time POS lever compensation method for an airborne LiDAR system according to claim 3, characterized in that, In step S3: In the Kalman filter of POS, the lever arm deviation state quantity is expanded. ;use Based on the current estimated attitude, calculate and output the corrected virtual antenna coordinates for subsequent point cloud calculations.

5. The real-time POS lever compensation method for an airborne LiDAR system according to claim 4, characterized in that, Step S3 includes: Modeling the actual and nominal positions of virtual antennas tiny offset between ; The attitude matrix estimated based on the current filter and contain The nominal geometric model predicts the baseline vector as ; virtual antenna nominal coordinates The virtual antenna predicted coordinates are compared with the IMU position predicted by the filter plus the arm vector after the current attitude and arm state transformation, and the difference also constitutes the observation. Using the estimated Based on the current estimated attitude, calculate the projection of the lever arm deviation onto the navigation frame. The corrected virtual antenna coordinates are output as follows: This is used for subsequent point cloud calculations.

6. The real-time POS lever compensation method for an airborne LiDAR system according to claim 5, characterized in that, In point cloud solving: Based on the virtual antenna coordinates, coordinate correction is performed, and a high-precision POS is output for laser point cloud settlement.

7. The real-time POS lever compensation method for an airborne LiDAR system according to claim 4 or 5, characterized in that, Also includes: S4, parameter adaptive; Will Feedback is sent to the geometric model to update the effective body coordinates of the virtual antenna, dynamically optimizing the model itself.

8. The real-time POS lever compensation method for an airborne LiDAR system according to claim 6, characterized in that, Step S4 also includes: updating the weighting coefficients; Weighting coefficient Slowly adaptive adjustments are made based on residual information.

Images