A method for extracting the relative angular rate of spatial non-cooperative targets based on point cloud data
By utilizing the periodic variation of point cloud quantity and Fourier transform, the relative angular rate of non-cooperative targets in space is extracted, solving the problem of computational complexity in existing technologies and achieving fast and accurate attitude angular rate estimation.
Patent Information
- Application Number
- CN202310948031.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-07-31
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-07-31
AI Technical Summary
Existing technologies involve large computational loads and complexities in estimating the relative attitude angular rate of non-cooperative targets in space, making it difficult to obtain attitude information quickly and accurately.
By utilizing the periodic variation of the number of point clouds during the rotation of a non-cooperative target in space, and combining the Fourier transform method, the time domain information is converted into the frequency domain, the relative angular rate information is extracted, and the relative attitude angular rate is obtained by normalizing the number of point clouds and calculating the maximum frequency component.
It enables the rapid and simple acquisition of the average relative attitude angular rate of non-cooperative targets in space, reduces the cumulative error in attitude estimation, and is suitable for targets with constant rotation or small fluctuations.
Smart Images

Figure CN117148312B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of satellite autonomous on-orbit servicing, and specifically relates to a method for extracting the relative angular rate of non-cooperative targets in space based on point cloud data. Background Technology
[0002] Currently, on-orbit servicing technologies for cooperative spacecraft are relatively mature. However, a significant portion of the targets serviced in orbit are non-cooperative space objects, such as space debris, faulty satellites, rocket upper stages, and non-cooperative spacecraft. These objects lack cooperative beacons for auxiliary measurements and may even be in an uncontrolled tumbling state, all of which significantly impact on-orbit servicing. Therefore, state monitoring of non-cooperative space objects is of great value for on-orbit servicing, especially the average relative angular rate information, which plays a crucial role in attitude estimation.
[0003] Reference 1, "Non-cooperative 3D Reconstruction of Targets Based on Point Cloud Pose Averaging [J], Space Control Technology and Application, 2020, 46(1): 7," discloses a drift correction method based on pose averaging. This method addresses the drift problem in the pose estimation process, such as... Figure 1 As shown, the pose to be optimized and the relative measurement are described by an undirected graph. The pose averaging problem of the point cloud is transformed into the problem of minimizing the relative pose error between nodes. The absolute value of the error is selected as the objective function, which can effectively eliminate the drift problem in the incremental pose estimation process. Reference 2 "Model Reconstruction and Pose Optimization of Spatial Non-cooperative Rotating Targets [J], Optics and Precision Engineering, 2019, 27(8): 9" discloses a method to reduce the cumulative pose error caused by the continuous estimation of pose between adjacent frames. The method uses a feature matching method based on GLAROT-3D global feature descriptor for closed loop detection. The method based on curvature is used for the extraction of key points.
[0004] The aforementioned literature eliminates the cumulative error of attitude estimation by using loop closure detection to obtain relatively accurate attitude information and relative attitude angular rate information. However, these methods inevitably require the registration or feature extraction of a large number of point cloud images, making the algorithms complex and computationally intensive. Summary of the Invention
[0005] To address the problem of relative attitude angular rate estimation in the relative pose estimation process of non-cooperative targets in space, this invention proposes a method for extracting the relative angular rate of non-cooperative targets in space based on point cloud data. This method utilizes the periodic variation of the number of point clouds caused by the rotation of the non-cooperative target, which contains the relative angular rate information of the target. By combining this with Fourier transform, the time-domain variation of the number of point clouds is transformed into a frequency-domain variation, thereby extracting the relative angular rate information from the frequency features, thus achieving the extraction of the relative angular rate information of non-cooperative targets in space.
[0006] The technical solution adopted by this invention to solve its technical problem is as follows:
[0007] A method for extracting the relative attitude angular rate of a non-cooperative spatial target based on point cloud data is characterized by utilizing the periodic variation of the number of point clouds caused by the rotation of the non-cooperative spatial target. This information contains the relative angular rate information of the non-cooperative spatial target. The number of point clouds is normalized according to the relationship between the relative distance to the non-cooperative spatial target and the number of point clouds. Then, the time-domain variation information of the number of point clouds is transformed into a frequency-domain variation using Fourier transform, and the relative angular rate information is extracted from the frequency features, thus achieving the extraction of the relative angular rate information of the non-cooperative spatial target. The point cloud number normalization method is shown below:
[0008]
[0009] Where s n and r n All of these can be calculated based on lidar measurement data.
[0010] Then, the function of the normalized point cloud number changing over time. Perform a Fast Fourier Transform to transform the normalized point cloud quantity function from the time domain to the frequency domain, and then find the frequency component f that has the greatest impact on the change in point cloud quantity in the frequency domain. m .
[0011] Finally, the average relative attitude angular rate ω of the non-cooperative target in space is calculated using the following formula. m
[0012] ω m =πf m (1-2)
[0013] From this, we can obtain the average relative attitude angular rate ω of the non-cooperative target in space. m .
[0014] Compared with the prior art, the beneficial effects of the present invention are as follows:
[0015] This method utilizes the periodic variation in the number of point clouds caused by the rotation of a non-cooperative target in space. It then uses Fourier transform to convert the time-domain variation of the point cloud number into a frequency-domain variation, thereby extracting relative angular rate information from the frequency characteristics. This method is relatively simple, computationally inexpensive, and can accurately and quickly obtain the average relative attitude angular rate of a non-cooperative target in space. It can reduce or eliminate accumulated errors in relative attitude estimation and is suitable for targets with constant or small fluctuations in rotational angular velocity. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the relative attitude drift situation in the prior art without attitude accumulation error correction, where the horizontal axis represents the point cloud data sequence collected sequentially as the target attitude changes, and the vertical axis represents the attitude estimation accumulation error.
[0017] Figure 2 This is a schematic diagram of a scenario where a spacecraft is tracking a non-cooperative target in space, and the two spacecraft form a near-circular satellite formation.
[0018] Figure 3 This is a schematic diagram illustrating a scenario where a tracking spacecraft follows a non-cooperative target in space, with the tracking spacecraft and the non-cooperative target forming a series of satellites.
[0019] Figure 4 A schematic diagram of the projection of a non-cooperative target in space. Detailed Implementation
[0020] The present invention will be further described below with reference to embodiments and accompanying drawings, but these should not be construed as limiting the scope of protection of the present invention.
[0021] For a non-cooperative target in space that has lost its power source, its attitude motion in space is only affected by disturbances such as gravitational gradient torque, aerodynamic torque, radiation torque, and magnetic disturbance torque. However, for the spin motion of a non-cooperative target in space, these disturbances have little effect on its attitude motion. Therefore, in the short term, the attitude motion of a non-cooperative target in space can be approximated as the non-cooperative target rotating at a certain constant angular velocity, that is, its attitude angular velocity is approximately a constant value.
[0022] To measure the average relative attitude angular rate of a non-cooperative space target, this invention proposes using a tracking spacecraft to serially follow or orbit the non-cooperative space target for attitude measurement. In this method, the tracking spacecraft and the non-cooperative space target form a serial formation or a near-circular formation, such as... Figure 2 and Figure 3As shown in the diagram, the lidar is mounted on the tracking spacecraft, enabling continuous acquisition of point cloud data of non-cooperative targets in space. Furthermore, the tracking spacecraft maintains tri-axis stability during its orbital flight, ensuring the lidar's optical axis always points towards the non-cooperative target. In this scenario, the angular velocity of the tracking spacecraft relative to the non-cooperative target can be approximated as a constant value.
[0023] Based on the above analysis, the attitude angular velocity of both the non-cooperative target and the tracking spacecraft can be approximated as constant over a short period of time. Therefore, the attitude angular rate of the non-cooperative target relative to the tracking spacecraft is also approximated as constant. The following describes a method for estimating this relative attitude angular rate using point cloud data acquired by lidar.
[0024] By using lidar to collect point cloud data of non-cooperative targets in space, point cloud data of the targets at different times can be obtained. Due to the attitude motion of the non-cooperative targets relative to the tracking spacecraft, the number of point cloud data collected each time changes in a regular manner (except for cases where the number of point cloud data hardly changes, such as when they are relatively stationary or when the non-cooperative targets rotate around the lidar optical axis). That is, the change in the number of point cloud data over time contains the relative angular rate information of the non-cooperative targets in space.
[0025] The following describes a method for obtaining the relative angular rate information of non-cooperative targets in space based on the temporal variation of point cloud quantity:
[0026] First, the number of point clouds collected is normalized. When observing a non-cooperative target in space from the same perspective, the number of observable point clouds changes as the distance between the lidar and the non-cooperative target changes. Since the distance between the lidar and the non-cooperative target changes continuously and periodically as the tracking spacecraft follows or orbits the target, normalization correction is required when calculating the relative attitude angular rate using the number of point clouds.
[0027] Assume the lidar is pointing towards the center of a non-cooperative target in space, and the lidar measurement coordinate system is O. S For XYZ, the projected area of a non-cooperative target along the lidar optical axis is equivalent to L*W, such as... Figure 4 As shown. Let s be the number of point clouds collected in the nth (n is a positive integer) collection. n At this time, the distance O between the non-cooperative target in space and the lidar is... S O S′ For r n The performance parameters of the lidar are: horizontal angular resolution α H The vertical angular resolution is α V When L < <r n And W < <r nThus, the number of point clouds s collected each time can be obtained. n for
[0028]
[0029] The number of point clouds s collected can be obtained from the above formula. n and Inversely proportional, therefore we can normalize the number of point cloud data as follows:
[0030]
[0031] in, r represents the normalized point cloud data volume over time. n It can be calculated from point cloud data obtained by lidar measurement, s n The number of points can be obtained by statistically analyzing the point cloud data measured by lidar. Where r... n The calculation method is as follows:
[0032] Let P be the coordinates of any point in the point cloud data obtained by the lidar in the nth measurement. i (x i y i z i ), where i is not greater than s n If r is a positive integer, then n The calculation formula is as follows:
[0033]
[0034] Then, after data normalization, a Fast Fourier Transform (FFT) is performed on the point cloud quantity information in the time domain. After the FFT, the point cloud quantity information is transformed from the time domain to the frequency domain, and the frequency component f that has the greatest impact on the change in point cloud quantity is identified in the frequency domain. m The unit is Hz.
[0035] Finally, based on the above maximum frequency component f m The average relative attitude angular rate of a non-cooperative target in space is calculated. From equation (1-3), it can be seen that the number of point clouds s... n Proportional to the projected area WL, for a non-cooperative spatial target undergoing relative attitude motion, the projected area after a 0° attitude rotation is the same as the projected area after a 180° attitude rotation, meaning the number of points in the point cloud remains unchanged. Therefore, the average relative attitude angular rate ω of a non-cooperative spatial target is... m for
[0036] ω m =πf m (1-6)
[0037] From this, we can obtain the average relative attitude angular rate ω of the non-cooperative target in space. m .
Claims
1. A method for extracting the relative attitude angular rate of a spatial non-cooperative target based on point cloud data, characterized in that, include: Step 1: Use a lidar to collect point cloud data of a non-cooperative target in space. The lidar is installed on a tracking spacecraft. The tracking spacecraft maintains three-axis stability while following or orbiting the non-cooperative target in space, and the optical axis of the lidar always points towards the non-cooperative target in space. The number of point clouds collected by the lidar in step 1 for the nth time It meets the following conditions: L<< ,W<< Where L is the length of the projected area of the non-cooperative target in the direction of the lidar optical axis when the lidar is pointing at the center of the non-cooperative target in space; W is the width of the projected area of the non-cooperative target in the direction of the lidar optical axis when the lidar is pointing at the center of the non-cooperative target in space. The distance between the lidar and the non-cooperative target in space during the nth data acquisition; This refers to the horizontal angular resolution of the lidar. This refers to the vertical angular resolution of the lidar. Let the coordinates of any point in the point cloud data obtained by the lidar in the nth measurement be... i is not greater than Given a positive integer, calculate the distance between the lidar and a non-cooperative target in space during the nth data acquisition. The formula is as follows: Step 2. Normalize the number of points based on the relationship between the relative distance to non-cooperative targets in space and the number of points, using the following formula: in, This is a function representing the normalized point cloud data volume over time. Step 3. Combining the Fourier transform method, the time-domain variation information of the number of point clouds is transformed into a frequency-domain variation, and then the relative angular rate information is extracted from the frequency features to realize the extraction of the relative angular rate information of non-cooperative targets in space.
2. The method for extracting the relative attitude angular rate of a spatial non-cooperative target based on point cloud data according to claim 1, characterized in that, Step 3 specifically involves: Step 3.1 Perform FFT (Fast Fourier Transform) on the number of points in the time domain, that is, convert the point cloud quantity information from the time domain to the frequency domain; Step 3.2 Find the frequency component that has the greatest impact on the change in the number of points in the point cloud from the point cloud quantity information in the frequency domain. ; Step 3.3 Calculate the average relative attitude angular rate of the non-cooperative target in space. The formula is as follows: 。
Citation Information
Patent Citations
Method for rapidly measuring relative pose of spatial non-cooperative slow-rotation target by laser radar
CN115963503A
