A sky survey method based on greedy algorithm considering satellite constellation avoidance

Abstract

The present invention relates to the field of satellite communication technology, and in particular to a sky survey observation method based on a greedy algorithm and taking satellite constellation avoidance into consideration, which can significantly reduce the pollution of satellite constellations to astronomical image data and thus improve the efficiency of sky survey observation. The method is characterized in that, according to observation conditions and orbital prediction information of satellite constellations, sky regions are divided, the number of satellite constellations in the field of view of each sky region and the length of the stripes they leave are calculated in real time, and a greedy algorithm is used to make the length of the stripes in the field of view of the telescope each time the shortest and the number of satellites the least, so as to reduce the probability of satellite constellations appearing in the field of view.

Description

Technical Field

[0001] The present invention relates to the field of satellite communication technology, and in particular to a sky survey observation method based on a greedy algorithm and taking satellite constellation avoidance into consideration, which can significantly reduce the pollution of satellite constellations to astronomical image data and thereby improve the efficiency of sky survey observation. Background Art

[0002] In the past four years, the number of satellites launched into Earth orbit has almost tripled. Taking the Starlink and OneWeb constellations, which consist of more than 40,000 individual satellites, as an example, the satellites in the corresponding projects have either been partially launched or are planned to be launched in the near future, which has led to a rapid increase in the number of low-orbit satellites. At the same time, astronomical observation scenes with large fields of view and long exposures are increasing. Many of these satellite constellations are highly reflective. When passing through the field of view of the telescope, they will leave bright "streaks" on the astronomical observation images. These "streaks" will contaminate the observation images and interfere with the precise measurement of astronomical targets.

[0003] In order to deal with the pollution of satellite constellations on astronomical observations, researchers are actively exploring feasible countermeasures. In the past, people often used sigma-clipping on multiple images to eliminate the influence of occasional noise (such as cosmic rays, meteors, etc.). This method is also applicable to the elimination of "streaks" of satellite constellations. However, deliberately converting a long-exposure image into multiple short-exposure images will not only increase the readout time, but also increase data storage. Other efforts are underway to mitigate the impact of satellite streaks on astronomical images. For example, satellite companies like SpaceX have been working to darken the exterior of satellites so that they are less conspicuous. However, even with the most effective dimming mitigation measures to date, satellites still appear bright on the camera and can cause effects such as nonlinear crosstalk or flicker. Summary of the invention

[0004] In response to the current problem of increasingly serious pollution of astronomical image data by satellite constellations due to the successive launch of large-scale low-orbit satellite plans, the present invention proposes a sky survey observation method based on a greedy algorithm that takes satellite constellation avoidance into account, which can effectively avoid low-orbit satellite constellations and significantly alleviate the impact of satellite constellations on astronomical image data, thereby providing safe and reliable support for subsequent satellite communication technologies.

[0005] The present invention is achieved by the following measures:

[0006] A sky survey observation method based on a greedy algorithm and taking satellite constellation avoidance into consideration is characterized in that the sky area is divided according to the observation conditions and the orbit prediction information of the satellite constellation, the number of satellite constellations in the field of view and the length of the stripes left by them in each sky area are calculated in real time, and the greedy algorithm is used to make the length of the stripes in the field of view of the telescope each time as short as possible and the number of satellites as small as possible, so as to reduce the probability of satellite constellations appearing in the field of view.

[0007] The present invention specifically comprises the following steps:

[0008] Step 1: Divide the sky area: When dividing the sky area, the effective field of view is defined as 10°×18°, the maximum distance from the moon is taken as 24 degrees, and the visible area of ​​the celestial sphere is divided into 10x18 grids;

[0009] Step 2: Obtain orbital prediction information of the satellite constellation: Obtain the orbital prediction information of the day through satellite constellation orbit calculation to determine the real-time position and visibility parameters of the satellite, including the time of the designated observation station at night, the visible time of the satellite, the visible time of the satellite from the observation station at night, and the time when the satellite is illuminated by the sun during the visible time, and obtain the right ascension, declination, altitude angle, azimuth angle and ID data information of different satellites at each moment;

[0010] Step 3: According to the orbit prediction information of the satellite constellation and the size of the telescope field of view, the number of satellite constellations falling in the telescope field of view and the length of the stripes passed by the satellite constellations in the field of view are calculated. The spherical angle distance of each satellite in the telescope field of view for the first and last appearance is recorded as the stripe length, and the number of satellites is recorded. The observation time of each field of view is regarded as a time step. In each time step, if the spherical angle distance of the right ascension and declination of the satellite constellation from the center coordinates of the field of view is less than half of the field of view, then the satellite is considered to be in the telescope field of view, and a data list is obtained with each field of view as the center;

[0011] Step 4: Traverse each field of view, obtain the stripe lengths in different fields of view within the current time step, and select the sky area that is closest to the current sky area and has the shortest stripe length.

[0012] In step 1 of the present invention, it is assumed that the sky area is divided into T i , where i∈{1,2,3,……,N}, represents the i-th sky area. The state of each sky area needs to be judged and recorded in different time steps. Let the time step be ,in represents the kth time step, and the target is ,in .

[0013] The stripe length in step 3 of the present invention is recorded as L i，k，Indicates the sky area T i At time step t k If the degree of stripe contamination within L i，k =0, indicating that there is no stripe effect in this area.

[0014] In the present invention, two points are set and The spherical distance formula is: (1), where α represents right ascension, β represents declination, and the sky area The center coordinates of , half of the field of view is ,Target In time step The coordinates of , Zetian District The number of targets in for:

[0015] (2),

[0016] In formula (2), 1 is the indicator function, which is 1 when the condition is met. The condition is that the spherical angle distance from the target to the center coordinate of the field of view is less than half of the field of view, then the target is considered to be in the field of view, otherwise it is 0. Vi is the right ascension and declination coordinates of the center of the ith field of view, recorded as , is half of the field of view; is the time step t k The right ascension and declination coordinates of the jth target are denoted as ;

[0017] Target O j Sky Zone T i The stripe length L j，i By target at first appearance time t first and last seen time t last Coordinate calculation:

[0018] (3),

[0019] Sky Zone T i Total stripe length L i （t k ） for:

[0020] (4),

[0021] At the current time step tk In this case, the area with the least number of targets and the shortest streak length is preferred. :

[0022] (5),

[0023] In the selection of the sky area for the next time step, the sky area with the shortest stripe length is selected first, and then the one closest to the current sky area is selected from the shortest sky area, that is, in the time step , select the area corresponding to the current sky The closest area of ​​the sky with the shortest streak length :

[0024] The entire observation path is , satisfying the following optimization objectives:

[0025] (7),

[0026] The present invention aims to minimize L, reduce the proportion of streak-contaminated images and the number of contaminated pixels during the survey process, and thus reduce the overall streak contamination.

[0027] The present invention avoids the pollution of astronomical images by satellite constellations by creating satellite constellation orbit forecasts, analyzing the visibility and density distribution characteristics of satellite constellations relative to observation stations, quantitatively evaluating the impact of satellite constellations on astronomical observations, and formulating avoidance observation strategies. Experiments have shown that the present invention can reduce the number of polluted images by 24.15% and the number of polluted pixels per image by 73.3% on average, thereby providing subsequent operational stability and reliability guarantees for satellite communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Attached Figure 1 21, 2024, is a schematic diagram of the observable sky area division of the celestial sphere on December 21, 2024 in an embodiment of the present invention. In the figure, the gray block area is the observable sky area, the gray dot represents the moon, and the blank area is the unobservable sky area.

[0029] Attached Figure 2 Schematic diagram of the WFOTA structure and field of view in an embodiment of the present invention.

[0030] Attached Figure 3 Schematic diagram of the observable sky area division of the celestial sphere on December 13, 2024 in an embodiment of the present invention.

[0031] Attached Figure 4 Schematic diagram of the observation sequence of the traditional scheduling sky survey in an embodiment of the present invention.

[0032] Attached Figure 5 Schematic diagram of the observation sequence of the observation strategy for avoiding satellite constellations in an embodiment of the present invention.

[0033] Attached Figure 6 is an image observed in an embodiment of the present invention, wherein Figure 6 In the figure, a is the observation image obtained by the traditional scheduling sky survey observation strategy, and b is the observation image obtained by avoiding the satellite constellation strategy.

[0034] Attached Figure 7 is an image observed in an embodiment of the present invention, wherein Figure 7 In the figure, a is the observation image obtained by the traditional scheduling sky survey observation strategy, and b is the observation image obtained by avoiding the satellite constellation strategy. DETAILED DESCRIPTION

[0035] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0036] Example

[0037] The telescope used in this example is WFOTA (Shandong University Xinjiang Muztagh Ata Small Photoelectric Array), and the effective field of view is defined as 10°×18°. In order to effectively divide the celestial sphere, the altitude angle is screened according to the astronomical observatory site. At the same time, considering the visibility of the target and the influence of the moon phase, the maximum distance from the moon is taken as 24 degrees to fully avoid the influence of the moon. The visible area of ​​the celestial sphere is divided into multiple grids of 10x18 size, and the movement speed of the right ascension axis and the declination axis are set according to the specific situation of the telescope. The schematic diagram of the division of the celestial area of ​​the entire celestial sphere is shown in the figure below. Figure 1 As shown, Figure 1 Schematic diagram of the sky area division on December 21, 2024.

[0038] Next, this example obtains the orbit forecast information for the day through satellite constellation orbit calculation. The main goal of the orbit calculation is to determine the real-time position and visibility parameters of the satellite, including the time of the designated observation site at night, the visible time of the satellite, the visible time of the satellite to the observation station at night, and the time when the satellite is illuminated by the sun during the visible time, and obtain data information such as the right ascension, declination, altitude angle, azimuth angle, and number ID of different satellites at each moment.

[0039] According to the orbit prediction information of the satellite constellation and the size of the telescope's field of view, the number of satellite constellations that fall within the telescope's field of view and the length of the stripes that the satellite constellations pass through in the field of view are calculated. The spherical angle distance that each satellite passes through when it first appears and last appears in the telescope's field of view is used as its stripe length, and the number of satellites is recorded. The observation duration of each field of view can be regarded as a time step. In each time step, the spherical angle distance of the right ascension and declination of the satellite constellation from the center coordinates of the field of view is less than half of the field of view. We believe that the satellite is within the telescope's field of view. A data list centered on each field of view is obtained, and each field of view is traversed to obtain the satellite distribution in different fields of view within the current time step (the number of satellites and id), the number of satellites is shown in formula (2).

[0040] The stripe length calculation is shown in formula (3). The coordinates of each satellite in the field of view are recorded, and the spherical angle distance between the first and last appearance is calculated as the stripes it passes through in the field of view. Then, each field of view can be traversed to obtain the stripe lengths in different fields of view within the current time step.

[0041] Specifically, in this example, the sky area division performs the following operations:

[0042] Assume that the sky is divided into T i , where i∈{1,2,3,…N} represents the ith sky region, and the state of each sky region needs to be judged and recorded in different time steps. Let the time step be t k , where k∈{1,2,3,…,K} represents the kth time step. The goal is O j , where j∈{1,2,3,…M}, and the stripe length is L i，k Indicates the sky area T i At time step t k The degree of stripe contamination within, if L i，k = 0, indicating that there is no stripe effect in the sky area, where two points A (α A , β A ) and B (α B , β B ) is:

[0043] (1),

[0044] Set up a sky area The center coordinates of , half of the field of view is , target O j At time step t k The coordinates of , Tianqu The number of targets in for:

[0045] (2),

[0046] Where 1 is the indicator function, which is 1 when the condition is met and 0 otherwise. Vi is the central right ascension and declination coordinates of the ith field of view, denoted as , is half of the field of view; At time step t k The right ascension and declination coordinates of the jth space debris are denoted as ;

[0047] Target O j In the Sky Zone T i Stripe length By target at first appearance time and last seen time Coordinate calculation:

[0048] (3),

[0049] Sky Region Total stripe length for:

[0050] (4),

[0051] At the current time step t k , prioritize the area T with the least number of targets and the shortest stripe length selected (t k ):

[0052] (5),

[0053] In the selection of the sky area for the next time step, the sky area with the shortest stripe length is selected first, and then the one closest to the current sky area is selected from the shortest sky area. k+1 , select the area T that matches the current area selected (t k ) The area T with the shortest fringe length selected (t k+1 ):

[0054] (6),

[0055] The entire observation path is P={T i1 , T i2 , ... T ik}, satisfying the following optimization goals:

[0056] (7),

[0057] The method proposed in this example aims to minimize L, reduce the proportion of streak-contaminated images and the number of contaminated pixels during the survey, and thus reduce the overall streak contamination.

[0058] The following is an analysis of the performance of this example:

[0059] To test the performance of the observation method described in this example, experiments were conducted using WFOTA at Shandong University in China. This strategy is also applicable to other sky survey telescopes. WFOTA consists of four 15 cm aperture large field optical telescopes, each with a field of view of about 100 square degrees. The four telescopes are mounted on two turntables, with two telescopes in each group. The spliced ​​field of view of each two telescopes mounted on a turntable is 10° x 18°. The total field of view of the four telescopes is 360 square degrees. The main task of WFOTA is to conduct large-scale sky surveys. The schematic diagram of WFOTA is shown in Figure 2 shown.

[0060] In order to make comparisons and ensure the same observation conditions, two groups of telescopes were used every night. On December 13, one group of telescopes A adopted the traditional observation strategy of scheduling sky surveys, and one group of telescopes B used the observation strategy of avoiding satellite constellations, and the observation results were compared.

[0061] Obtain the TLE data of the satellite constellations Starlink and OneWeb on December 13, 2024, perform orbit prediction calculations based on the environmental conditions and equipment conditions of WFOTA, and obtain the visible time, right ascension, declination, altitude angle, azimuth angle, etc. of the satellite constellation. According to the requirements of actual observation, the influence of target visibility and moon phase are also considered. We take 75 seconds as a time step, that is, we shoot continuously for 75 seconds in each direction of the sky area, expose once every 10 seconds, and the data readout time is 0.98 seconds, and obtain the division of the sky area pointed by the telescope, such as Figure 3 shown.

[0062] The observation order of the traditional scheduled sky survey observation strategy is defined by the following description: among the current 8 adjacent sky areas, start from the right side as the first one, and search for unobserved sky areas in a counterclockwise direction, right-upper-left-lower-right-upper-left-lower-left-lower-right. If found, observe directly. If not found, expand to the second circle of declination until the end.

[0063] According to the above observation order of the sky area, assuming that the number of the sky area is i, there are N sky areas in total. Pointing to the sky area can be done with T i Represented by , where i∈{1,2,3,…N}, the observed target can be represented by Oj, where j∈{1,2,3,…M}, M is the total number of targets.

[0064] Set the direction to the sky area T i The number of targets contained in is n i , target O j Whether it falls into T i It can be judged by the following expression:

[0065] (8),

[0066] in , if the target Falling in the sky ; otherwise 0, if , then record the target and set the target In the sky area The length of the stripes in , Zetian District The total stripe length of all targets in is:

[0067] (9),

[0068] in: is the target exist The stripe length, Used to determine whether the target belongs to the sky area , the total stripe length of the polluted image is:

[0069] (10),

[0070] like Figure 4 Shown is a schematic diagram of the observation sequence of the traditional scheduled sky survey on December 13, 2024 (13:00:00-14:48:45).

[0071] In this example, the observation sequence diagram of the observation strategy to avoid satellite constellations on December 13, 2024 (13:00:00-14:48:45) is as follows: Figure 5 shown.

[0072] Through the method of manual marking and calculation, first mark the pixel coordinates of the beginning of each stripe in the image, then mark the pixel coordinates of the end of each stripe, and finally calculate the distance between the two pixel coordinates as the length of the satellite stripe. The observation results on December 13, 2024 show that in the observation images obtained by the traditional scheduling sky survey observation strategy, the number of images contaminated by stripes accounted for 37.69% (349 / 926), and the average stripe length was 577.12 pixels / frame; in the observation images obtained by avoiding the satellite constellation strategy, the number of images contaminated by stripes accounted for 13.54% (114 / 842), and the average stripe length was 154.10 pixels / frame. In the observation images obtained using the avoidance strategy, the proportion of contaminated images was reduced by 24.15%, and the average stripe length in the observation images was reduced by 423.02 pixels / frame. The comparison results of some observations on December 13, 2024 are as follows Figure 6 , Figure 7 shown.

[0073] It can be seen that with the rapid increase of satellite constellations in the existing technology, the observation strategy of astronomical survey observation has become more and more complicated. This example analyzes the visibility and density distribution of satellite constellations relative to ground observation stations, and scientifically, quantitatively and systematically evaluates the impact of satellite constellations on astronomical observations based on reasonable sky area division. The proportion of the number of contaminated images and the number of contaminated pixels are used to reduce the probability of satellite constellations appearing in the telescope field of view, and obtain an observation strategy to avoid satellite constellations. Experiments have shown that although 9% of the observation time is lost to avoid constellations by using the observation strategy to avoid satellite constellations, this method can reduce the number of contaminated images by 24.15%, and the average number of contaminated pixels per image is reduced by 73.3%. The avoidance strategy proposed in this example is worthwhile. The observation strategy to avoid satellite constellations can effectively reduce the number of contaminated images in the observed images, significantly shorten the length of satellite stripes in the observation, thereby greatly reducing the impact of the average pixel loss of each observation, and greatly alleviating the impact of satellite constellations on astronomical survey observations.

Claims

1. A sky survey observation method based on a greedy algorithm and considering satellite constellation avoidance, characterized in that: According to the observation conditions and the orbit prediction information of the satellite constellation, the sky area is divided, and the number of satellite constellations in the field of view and the length of the stripes left by them in each sky area are calculated in real time. The greedy algorithm is used to make the length of the stripes in the field of view of the telescope the shortest each time and the number of satellites the least, so as to reduce the probability of satellite constellations appearing in the field of view; The specific steps include: Step 1: Divide the sky area: When dividing the sky area, the effective field of view is defined as 10°×18°, the maximum distance from the moon is taken as 24 degrees, and the visible area of ​​the celestial sphere is divided into 10x18 grids; Step 2: Obtain orbital prediction information of the satellite constellation: Obtain the orbital prediction information of the day through satellite constellation orbit calculation to determine the real-time position and visibility parameters of the satellite, including the time of the designated observation station at night, the visible time of the satellite, the visible time of the satellite from the observation station at night, and the time when the satellite is illuminated by the sun during the visible time, and obtain the right ascension, declination, altitude angle, azimuth angle and ID data information of different satellites at each moment; Step 3: According to the orbit prediction information of the satellite constellation and the size of the telescope field of view, the number of satellite constellations falling in the telescope field of view and the length of the stripes passed by the satellite constellations in the field of view are calculated. The spherical angle distance of each satellite in the telescope field of view for the first and last appearance is recorded as the stripe length, and the number of satellites is recorded. The observation time of each field of view is regarded as a time step. In each time step, if the spherical angle distance of the right ascension and declination of the satellite constellation from the center coordinates of the field of view is less than half of the field of view, then the satellite is considered to be in the telescope field of view, and a data list is obtained with each field of view as the center; Step 4: Traverse each field of view, obtain the stripe lengths in different fields of view within the current time step, and select the sky area that is closest to the current sky area and has the shortest stripe length.

2. The method for sky survey observation based on greedy algorithm and considering satellite constellation avoidance according to claim 1 is characterized in that: In step 1, assume that the sky area is divided into T i , where i∈{1,2,3,…N} represents the ith sky region. The state of each sky region needs to be judged and recorded in different time steps. Let the time step be t k , where k∈{1,2,3,…K}, represents the kth time step, and the target is O j , where j∈{1,2,3,…M}.

3. The method for sky survey observation based on greedy algorithm and considering satellite constellation avoidance according to claim 1, characterized in that: The stripe length in step 3 is recorded as L i，k， Indicates the sky area T i At time step t k If the degree of stripe contamination within L i，k =0, indicating that there is no stripe effect in the sky area.

4. The method for sky survey observation based on greedy algorithm and considering satellite constellation avoidance according to claim 2 is characterized in that: Set two points A (α A , β A ) and B (α B , β B） The spherical distance formula is: （1）， Among them, α represents right ascension, β represents declination, and the sky area T i The center coordinates of i , δ i ), half of the field of view is r v , target O j At time step t k The coordinates of (α j (t k ),δ j (t k )), Zetian District T i The number of targets in N i （t k ） for: （2）， In formula (2), 1 is the indicator function, which is 1 when the condition is met, otherwise it is 0. The condition is: if the spherical angle distance from the target to the center coordinate of the field of view is less than half of the field of view, the target is considered to be in the field of view. Vi is the right ascension and declination coordinates of the center of the ith field of view, denoted as (α i , δ i ), r v is half of the field of view; D j (t k ) is the time step t k The right ascension and declination coordinates of the jth target are denoted as ; Target O j In the Sky Zone T i The stripe length L j，i The target first appears at time t first and the last appearance time t last Coordinate calculation: （3）， Sky Zone T i Total stripe length for: （4）， At the current time step , giving priority to the area with the least number of targets and the shortest streak length : （5）， In the selection of the sky area for the next time step, the sky area with the shortest stripe length is selected first, and then the one closest to the current sky area is selected from the shortest sky area, that is, in the time step t k+1 , select the area corresponding to the current sky The closest area of ​​the sky with the shortest streak length : （6）， The entire observation path is , meeting the following optimization goals: （7）， Designed to minimize , reducing the proportion of streak-contaminated images and the number of contaminated pixels during the survey, thereby reducing overall streak contamination.

Images