Space traffic information sensing method based on star sensor network
By installing a star sensor on the space-based platform, a star sensor network is formed, which solves the problem that the existing technology is difficult to perceive long-distance space targets, and effectively monitors space targets and establishes a cataloging library, which improves the perception of space traffic information.
Patent Information
- Application Number
- CN202411294132.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2025-06-06
AI Technical Summary
The prior art is difficult to achieve effective perception and monitoring of long-distance space targets, especially due to the volume and power consumption limitations of space-based radars, it is impossible to detect long-distance space targets.
By installing star sensors on multiple space-based platforms, a star sensor network is formed, and optical measurements are used to identify space targets, and the orbit determination and monitoring of space targets is achieved through image processing and information fusion.
It has achieved effective perception and monitoring of long-distance space targets, improved my country's space traffic information perception capabilities, established a fully autonomous space target catalog library, and ensured the in-orbit safety of the spacecraft.
Smart Images
Figure CN120110482A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a space traffic information perception method based on a star sensor network. It involves using a large number of satellites equipped with star sensors to form a space-based optical observation network, processing and information fusion of data obtained by each star sensor node, and realizing wide-area and rapid space traffic information perception. It belongs to the field of space situational awareness. Background Art
[0002] With the rapid development of human space technology, the number of spacecraft entering outer space is increasing day by day, and countries around the world have successively proposed low-orbit giant constellation plans. In recent years, the number of space targets has shown an exponential increase.
[0004] Space-based platforms can be used to achieve wide-area, high-performance space traffic perception. The detection equipment used mainly includes two categories: space-based radars and star sensors. Space-based radars are currently unable to detect distant space targets due to their size and power consumption. Star sensors are optical measurement devices that take space pictures in a specified direction. The pictures will contain actively luminous stars and space targets that reflect sunlight. Space targets are identified through image processing, and the direction of the space targets relative to the star sensors is obtained. Multiple measurements can be used to determine the orbit and other information of the space targets, and monitor the space targets. Star sensors have the advantages of small size, power consumption, high precision, and relatively low cost. By carrying star sensors on multiple space-based platforms to form a star sensor network, a space-based space target monitoring network can be established to enhance my country's space traffic information perception capabilities and establish a fully autonomous space target catalog library in my country. Summary of the invention
[0005] In order to realize the timely perception of space traffic information and ensure the on-orbit safety of our spacecraft, the present invention provides a space traffic information perception method based on a star sensor network. The present invention comprises two parts, a space end and a ground end. The space end refers to the star sensor network. The star sensor uses an onboard computer to complete image acquisition and processing on orbit. The ground end refers to a ground data processing system, which analyzes the measurement data of the star sensor network. The space end transmits data to the ground end through a satellite measurement and control network.
[0006] The present invention provides the following technical solutions:
[0007] S1: Based on satellite networking, each satellite in the network is equipped with a star sensor to form a space traffic sensing network;
[0008] S2: The star sensor uses the onboard computer of the device to complete image processing and astronomical positioning, identify space targets, obtain the right ascension and declination measurements of space targets, and transmit the right ascension and declination measurements to the satellite's star service computer;
[0009] S3: The satellite's measurement and control system transmits the measurement arc segment back to the ground measurement and control station at the set time interval, and the ground measurement and control station transmits the measurement arc segment to the ground data processing system;
[0010] S4: The ground data processing system processes the measurement arc according to the processes of initial orbit determination, data association, catalog library establishment / update, maneuver detection, collision warning, etc., to achieve wide-area and rapid perception of space traffic information.
[0011] In the present invention, the image acquisition and processing process at the space end includes the following steps:
[0012] (1) Obtain the original image, preprocess the original image, minimize the impact of noise, improve image quality, and extract all targets in the image based on feature information;
[0013] (2) Perform star map matching, select stars within the visible range based on star sensor parameters and coarse-precision optical axis pointing, and match the targets in the image with the stars;
[0014] (3) Space target recognition and astronomical attitude determination: non-stellar targets are initially screened out based on the star map matching results. The inertial attitude of the star sensor is determined using the star map matching results. Based on the attitude information and orbit information of the star sensor, multiple frames of images are combined to further accurately identify the space targets in the image.
[0015] (4) Determine the target orientation. Use the two-dimensional distribution of matching stars in the image and the attitude of the star sensor to construct a one-to-one mapping relationship between the two-dimensional pixel coordinates in the image and the inertial orientation. Calculate the inertial orientation of the space target relative to the star sensor based on the two-dimensional coordinates of the space target in the image.
[0016] In the present invention, the data processing process of the ground terminal includes the following steps:
[0017] (1) Data preprocessing: Time synchronization of the orbit data of each star sensor with the space target measurement data, and elimination of outliers in the measurement arc;
[0018] (2) Extract the historical space target catalog library, which contains the orbital information of each cataloged target. It is assumed that all space targets are maneuvering. The historical orbital information of the space targets is associated with the current batch of measurement arcs, and the orbital information of the successfully associated targets is updated;
[0019] (3) extracting the measured arcs that failed to be associated in step (2), assuming that these arcs belong to the space targets that are undergoing maneuvers, performing arc association under the maneuvering assumption, and determining the maneuvering events. The orbit information of the maneuvering targets that are successfully associated is updated;
[0020] (4) extracting the measured arcs that could not be associated in step (2) and step (3), assuming that these arcs belong to uncataloged new targets, recording these arcs as UCTs, correlating the UCTs, determining the orbit of the new target, and writing the new target into the catalog library;
[0021] (5) Using the updated space target catalog, calculate the collision probability between our spacecraft and all space targets in the next three days. If the collision probability reaches the danger threshold, generate a collision warning report and design a collision avoidance strategy. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] Figure 1 This is the architecture diagram of the space traffic information perception system based on star sensor networking.
[0023] Figure 2 It is the data processing flow chart of the space end.
[0024] Figure 3 This is the data processing flow chart of the ground side. DETAILED DESCRIPTION
[0025] The technical scheme and process of the present invention will be described in detail and completely below with reference to the accompanying drawings.
[0026] The present invention discloses a method for sensing space traffic information based on a star sensor network. The overall architecture of the system is as follows: Figure 1 As shown in the figure, it consists of three parts: space end, ground end and user. On the space end, multiple artificial earth satellites form a regional or global network. Each satellite is equipped with a star sensor, thus forming a star sensor network. The star sensor takes space images in real time and dynamically. The original image is processed by the onboard computer of the star sensor itself to obtain the space target measurement data. Only the space target measurement data, not the original image, is transmitted to the star service computer. The star service computer packages the satellite orbit data and space target measurement data of the same time period, and transmits them to the ground control station through the measurement and control system. The ground control station transmits the data to the ground data processing center through optical cables. The data processing center analyzes and mines the space target measurement data, generates data products including space target orbit, collision probability and collision avoidance strategy, and provides them to users.
[0027] like Figure 2 As shown, the present invention provides a data processing flow at the space end, comprising the following steps:
[0028] (1) Image acquisition: Preprocess the original image, use stray light elimination measures and grayscale threshold to filter the background stray light, use threshold segmentation to reduce the impact of background brightness on the accuracy of star point centroid extraction, use low-pass filter template to suppress image noise, use median filter to remove starry sky background noise, and use Gaussian smoothing filter to suppress high-frequency noise;
[0029] (2) Feature extraction and centroid positioning: Point targets and trailing targets in an image have certain feature information. The point targets are extracted using the corner point extraction method, and the trailing targets are extracted using the edge detection method. The contour information of the target in the image, the grayscale information of the pixels occupied, etc. are obtained to determine the centroid position of the point target.
[0030] (3) Star map matching: The targets in the image mainly include stars and space targets. Space targets can be identified through the "subtraction" idea. The stars in the background are removed by using the star map matching method. The remaining features in the image are considered to be potential space targets. A navigation star library is constructed, and the triangle matching method is used to match the navigation star library with the star targets in the image.
[0031] (4) Target identification and astronomical attitude determination: In step (3), the star targets have been identified through star map matching. The remaining targets are potential space targets. The imaging configuration of the trailing target is relatively close to the straight line feature. The trailing target is detected based on the straight line feature, the endpoint of the trailing target is calculated, and the target motion direction is calculated. At the same time, the inertial attitude of the star sensor is determined using the star map matching results. According to the attitude information and orbit information of the star sensor, inter-frame target matching is performed, and the corresponding relationship between the targets in adjacent frames is established to complete the target classification;
[0032] (5) Calculate the target orientation: Construct a camera distortion model to obtain the mutual conversion relationship between the ideal image coordinates and the real image coordinates, determine the precise orientation of the space target in the star sensor imaging coordinate system, and use the astronomical attitude determination result in step (4) to convert the orientation of the space target relative to the star sensor into the inertial system.
[0033] like Figure 3 As shown, the present invention provides a data processing flow at the ground end, comprising the following steps:
[0034] (1) Data preprocessing: The data packets transmitted from the space end to the ground end mainly contain the orbit data of the star sensor and the measurement arc of the space target. The orbit data and the space target measurement data are usually not synchronized. The time of the space target measurement data is kept unchanged, and the original orbit data is used to calculate the position of the space target at the time of measurement by polynomial interpolation. A quadratic curve fitting is performed on a single measurement arc. Outliers in the measurement arc are manifested as excessively large fitting residuals, so outliers are eliminated.
[0035] (2) Extract the historical space target catalog library, predict the historical orbit of each cataloged target according to the orbital dynamics equation to the time of the current arc segment to be associated, use the orbit prediction result of the target and the position of the star sensor to calculate the measurement reference value, and obtain the residual between the measurement reference value and the actual measurement. If the residual is less than the set critical value, it is considered that the arc segment belongs to the space target. Perform the above operation on all measured arc segments, and then filter out the arc segments belonging to different space targets. If the association fails, continue the operation according to step (3). Assuming that a certain target is associated with several arc segments, use these arc segments to update the orbit information of the target in the catalog library;
[0036] (3) The arc segment association process in step (2) predicts the orbit of the space target according to the free orbit dynamics equation. However, in actual situations, spacecraft often undergo orbital maneuvers, and the orbit prediction results will have a large deviation from the actual values, resulting in association failure. This step targets the measured arc segments that failed to be associated in step (2). It is assumed that these arc segments belong to the space targets that have undergone maneuvers. For cooperative spacecraft, the orbit prediction results are corrected after obtaining the maneuvering information, and then association is performed. For non-cooperative spacecraft, arc segment association and maneuvering detection are performed synchronously based on the orbital reachable domain. After the arc segment association is completed, multiple arc segments are used to update the orbit information of the maneuvering target in the catalog library.
[0037] (4) Steps (2) and (3) only consider catalogued space targets. However, due to the launch of new spacecraft and the newly generated space debris, the measurement arcs belonging to the new targets cannot be associated in steps (2) and (3). The measurement arcs that cannot be associated in steps (2) and (3) are extracted. It is assumed that these arcs belong to uncatalogued new targets. These arcs are recorded as UCTs and the UCTs are correlated. If there are 4 or more UCTs correlated with each other, it is considered that a new target has been successfully catalogued. The orbit of the new target is calculated and the new target is written into the catalog library.
[0038] (5) Based on the updated catalog of space targets, screen out space targets whose orbital semi-major axis, inclination and right ascension of ascending node are close to those of our spacecraft. Use the analytical orbit predictor to determine whether these space targets are likely to rendezvous with our spacecraft within the next three days. For space targets that are likely to rendezvous, further use the numerical orbit predictor to make high-precision orbit predictions, predict orbit uncertainties, and calculate collision probabilities. If the collision probability reaches the danger threshold, generate a collision warning report and design a collision avoidance strategy.
Claims
1. A space traffic information perception method based on a star sensor network, characterized in that: It consists of two parts: the space end and the ground end. The space end is based on satellite networking. Each sub-satellite in the network is equipped with a star sensor, forming a space traffic perception network. The star sensor uses the onboard computer to complete image processing and astronomical positioning, identify space targets, obtain the right ascension and declination measurements of space targets, and transmit the right ascension and declination measurements to the satellite's space service computer. The satellite's measurement and control system transmits the measurement arc segment back to the ground at the set time interval. Based on the acquired measurement arcs, the data processing center on the ground carries out data processing tasks such as initial orbit determination, data association, catalog library establishment / update, maneuver detection, collision warning, etc., to achieve wide-area and rapid space traffic information perception, and provide users with space traffic information perception data products.
2. A star sensor on-orbit data processing process for space traffic information perception, characterized in that: (1) Obtain the original image, preprocess the original image, minimize the impact of noise, improve image quality, and extract all targets in the image based on feature information; (2) Perform star map matching, select stars within the visible range based on star sensor parameters and coarse-precision optical axis pointing, and match the targets in the image with the stars; (3) Space target recognition and astronomical attitude determination: non-stellar targets are initially screened out based on the star map matching results. The inertial attitude of the star sensor is determined using the star map matching results. Based on the attitude information and orbit information of the star sensor, multiple frames of images are combined to further accurately identify the space targets in the image. (4) Determine the target orientation. Use the two-dimensional distribution of matching stars in the image and the attitude of the star sensor to construct a one-to-one mapping relationship between the two-dimensional pixel coordinates in the image and the inertial orientation. Calculate the inertial orientation of the space target relative to the star sensor based on the two-dimensional coordinates of the space target in the image.
3. A ground data processing process for space traffic information perception, characterized in that: (1) Data preprocessing: Time synchronization of the orbit data of each star sensor with the space target measurement data, and elimination of outliers in the measurement arc; (2) Extract the historical space target catalog library, which contains the orbital information of each cataloged target. It is assumed that all space targets are maneuvering. The historical orbital information of the space targets is associated with the current batch of measurement arcs, and the orbital information of the successfully associated targets is updated; (3) extracting the measured arcs that failed to be associated in step (2), assuming that these arcs belong to the space targets that are undergoing maneuvers, performing arc association under the maneuvering assumption, and determining the maneuvering events. The orbit information of the maneuvering targets that are successfully associated is updated; (4) extracting the measured arcs that could not be associated in step (2) and step (3), assuming that these arcs belong to uncataloged new targets, recording these arcs as UCTs, correlating the UCTs, determining the orbit of the new target, and writing the new target into the catalog library; (5) Using the updated space target catalog, calculate the collision probability between our spacecraft and all space targets in the next three days. If the collision probability reaches the danger threshold, generate a collision warning report and design a collision avoidance strategy.