A fast prediction algorithm for optical remote sensing satellite imaging window in canyon environment

Abstract

The present invention introduces a fast prediction algorithm for optical remote sensing satellite imaging windows in a canyon environment, and the specific steps are: S1: constructing a visibility model of remote sensing satellites and ground target points; S2: using visible window semi-analysis for fast calculation; S3: constructing a terrain occlusion analysis model; S4: performance test analysis. The fast prediction algorithm for optical remote sensing satellite imaging windows in a canyon environment of the present invention takes into account the influence of terrain features such as canyons on the prediction of remote sensing satellite imaging windows, realizes accurate calculation of visible time windows, provides more time windows for users to perform tasks, and further provides strong support and guarantee for special users to perform tasks.

Description

Technical Field

[0001] The invention relates to the technical field of satellite remote sensing imaging, and in particular to a fast prediction algorithm for an optical remote sensing satellite imaging window in a canyon environment. Background Art

[0002] Remote sensing satellites are space resources that perform ground observation tasks. They have a large observation range and can obtain high-quality surface images. They have important applications in weather forecasting, disaster monitoring, and military reconnaissance. During the observation of ground target points by remote sensing satellites, the relative position of the satellite and the ground target point changes over time. Only when the ground target is within the visible range of the pitch angle and roll direction of the satellite's field of view, the remote sensing satellite is considered to be visible to the ground target point. When the satellite passes by the ground target, the time range in which the target can be observed under the field of view constraint is called the visible time window. It is the basis for the satellite to perform observation tasks on ground targets. Any observation task must be performed within the visible window, so it is of great significance to accurately calculate the time window. The tracking propagation method is a traditional algorithm for solving the satellite visible window. It uses a very small simulation step size Δt to predict the satellite orbit and target point, and judges whether the satellite is visible to the target point in each step of the simulation, and simulates the satellite's visible interval. This method is simple and the calculation results are accurate, but the efficiency is very low. In addition, in the analysis of existing research results, most of them only require the input of the coordinate information of ground target points and areas, without considering terrain features such as canyons. In plain areas, it has little impact on the visible time window of optical remote sensing satellites, but in areas with complex terrain, tall mountains will block part of the visible range of optical remote sensing satellites. If the influence of the canyon environment is ignored, the predicted optical remote sensing satellite imaging window will be significantly more than the actual imaging window, which is not conducive to the planning and execution of subsequent specific user tasks. Summary of the invention

[0003] The purpose of the present invention is to overcome the deficiencies of the prior art and provide a fast prediction algorithm for optical remote sensing satellite imaging windows in a canyon environment to achieve fast and accurate calculation of visible time windows.

[0004] The technical solution adopted by the present invention is: a fast prediction algorithm for optical remote sensing satellite imaging window in a canyon environment, the specific steps are:

[0005] S1: Construct visibility model between remote sensing satellite and ground target points

[0006] Define the maximum geocentric angle corresponding to the minimum observation angle of remote sensing satellite earth observation as d σ ;

[0007] At a certain time t, the latitude and longitude of the ground target point is (λ t , ), the longitude and latitude of the satellite subsatellite point is (λ s , ), and its calculation formula is:

[0008]

[0009] The calculation formula of the geocentric angle between the satellite sub-satellite point and the ground target point is:

[0010]

[0011] If d t-s ≤d σ , then at time t, the ground target point is within the instantaneous coverage area of ​​the satellite, otherwise, it is not within the coverage area; i is the orbital plane inclination, ω is the orbital perigee angular distance, Ω 0 is the right ascension of the ascending node, and f is the true anomaly;

[0012] S2: Fast calculation using visible window semi-analysis

[0013] First determine whether the ground target is on the same side of the earth as the satellite, and then determine whether it is within the visible range of the satellite; use a coarse-precise variable step search, use a large step to preliminarily determine the range of the start and end times, and then use a binary search to accurately calculate the start and end times, and then obtain the visible time window of the regional target;

[0014] S3: Constructing terrain occlusion analysis model

[0015] When a ground target point is located in the coverage area of ​​a remote sensing satellite at a certain moment, the visibility analysis method based on elevation angle and azimuth angle is used to analyze the impact of the surrounding high mountain terrain on the coverage area. If the terrain blocks the remote sensing satellite's line of sight, it means that the ground target point is not in the coverage area of ​​the remote sensing satellite at that moment.

[0016] S4: Performance Test Analysis

[0017] The accuracy and speed of the prediction algorithm of the present invention are verified through simulation examples, and different scenarios are distinguished for test evaluation;

[0018] Scenario 1: Without considering terrain features, the accuracy is compared with the commercial software STK, and the speed is compared with the classic tracking propagation method;

[0019] Scenario 2: Compare the accuracy and speed of the algorithm used in Scenario 1 with those in Scenario 1, taking terrain features into account.

[0020] Specifically, in the step S2, in the rough calculation, the discrete time step is relatively large, the purpose is to quickly determine the approximate time interval of the visible start time and the visible end time. The specific calculation steps are as follows:

[0021] a. Knowing the start and end time of the observation task, divide the simulation period into several sub-intervals. Let M 0 It is the point where the satellite passes through the orbital ascending node every time, and the angle between the ascending node and the ascending node is equal to a certain set value; from the start time to the first time the satellite passes through M 0 For the first paragraph, use (tS 1 ,tE 1 ) indicates that the satellite passes through M for the first time. 0 By the second time 0 For the second paragraph, use (tS 2 ,tE 2 ) means, divided in sequence, the last time through M 0 The last segment is when the simulation ends.

[0022] b. Discretize the entire process of the remote sensing satellite orbit into n equal parts, that is, n+1 discrete points, and record the corresponding time series as t i , i=1,2,...,n+1; According to the TLE file of the remote sensing satellite, the SGP4 model is used to predict the satellite orbit, and the satellite position and velocity information is quickly obtained through Lagrange interpolation;

[0023] c. Calculate each discrete time point t i Are the satellite and the ground point visible? If they are within the visible range, the satellite is considered to be at the discrete time point t i Visible to the target point;

[0024] d、t i It can be seen that t i-1 When it is not visible, the visible window starts at t i-1 to i Between, when t i-1 It can be seen that t i When it is not visible, the visible window ends at t i-1 to i between;

[0025] Through rough calculation, the approximate interval between the start and end time of the visible time window has been calculated. Using binary search, the precise visible time window can be efficiently calculated. The specific calculation steps are as follows:

[0026] e. If the visible window starts at the time interval (t 0 , t n ), calculate the intermediate time t m The visibility of the satellite to the ground point; if t m It can be seen that its value is assigned to t n If t m If it is not visible, assign its value to t 0 ;

[0027] f. Repeat step e until t n -t 0 <ε, ε is the given error limit;

[0028] g. Get the visible start time t, that is

[0029] The visible start time is thus obtained, and the calculation idea of ​​the visible end time is the same as that of the visible start time.

[0030] Specifically, in S3, in order to determine whether different terrains block the remote sensing satellite's line of sight, terrain elevation data under different environments is obtained through a DEM that can reflect the terrain elevation, and is read into the simulation software as the data basis for subsequent analysis; the specific steps are:

[0031] (1) Set up equally spaced sampling points within a range of 100-1000 meters in the direction of line of sight projection;

[0032] (2) Determine the longitude and latitude positions of each sampling point by transforming the coordinate system between the station center coordinate system and the earth coordinate system (λ sample , );

[0033] (3) Using the discrete point elevation information provided by the original DEM data of the mountain, the actual terrain elevation H at each sampling point is obtained through the bilinear interpolation fitting method. sample ;

[0034] (4) Compare the actual terrain of the sampling point with the elevation angle E of the ground target point l_sample The elevation angle E between the remote sensing satellite and the ground target point l The size of E l_sample <E l , then the ground target can be observed by the remote sensing satellite, otherwise the ground target cannot be observed by the remote sensing satellite.

[0035] Due to the adoption of the above-mentioned technical solution, the present invention has the following advantages:

[0036] The optical remote sensing satellite imaging window rapid prediction algorithm under the canyon environment of the present invention takes into account the influence of terrain features such as canyons on the remote sensing satellite imaging window prediction, realizes the accurate calculation of the visible time window, provides more time windows for users to perform tasks, and further provides strong support and guarantee for special users to perform tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is the algorithm flow chart of this patent. DETAILED DESCRIPTION

[0038] The present invention will be further explained below in conjunction with the accompanying drawings and embodiments, which should not be used to limit the protection scope of the present invention. The purpose of disclosing the present invention is to protect all technical improvements within the scope of the present invention.

[0039] Combined with Figure 1 The optical remote sensing satellite imaging window rapid prediction algorithm in a canyon environment is shown in the figure, and the specific steps are as follows:

[0040] S1: Construct visibility model between remote sensing satellite and ground target points

[0041] Define the maximum geocentric angle corresponding to the minimum observation angle of remote sensing satellite earth observation as d σ ;

[0042] At a certain time t, the latitude and longitude of the ground target point is (λ t , ), the longitude and latitude of the satellite subsatellite point is (λ s , ), and its calculation formula is:

[0043]

[0044] The calculation formula of the geocentric angle between the satellite sub-satellite point and the ground target point is:

[0045]

[0046] If d t-s ≤d σ , then at time t, the ground target point is within the instantaneous coverage area of ​​the satellite, otherwise, it is not within the coverage area; i is the orbital plane inclination, ω is the orbital perigee angular distance, Ω 0 is the right ascension of the ascending node, and f is the true anomaly.

[0047] S2: Fast calculation using visible window semi-analysis

[0048] First determine whether the ground target is on the same side of the earth as the satellite, and then determine whether it is within the visible range of the satellite; use a coarse-precise variable step search, use a large step size to preliminarily determine the range of the start and end times, and then use a binary search to accurately calculate the start and end times, and then obtain the visible time window of the regional target.

[0049] In the rough calculation, the discrete time step is large, the purpose is to quickly determine the approximate time interval of the visible start time and the visible end time. The specific calculation steps are as follows:

[0050] a. Knowing the start and end time of the observation task, divide the simulation period into several sub-intervals. Let M 0 It is the point where the satellite passes through the orbital ascending node every time, and the angle between the ascending node and the ascending node is equal to a certain set value; from the start time to the first time the satellite passes through M0 For the first paragraph, use (tS 1 ,tE 1 ) indicates that the satellite passes through M for the first time. 0 By the second time 0 For the second paragraph, use (tS 2 ,tE 2 ) means, divided in sequence, the last time through M 0 The last segment is when the simulation ends.

[0051] b. Discretize the entire process of the remote sensing satellite orbit into n equal parts, that is, n+1 discrete points, and record the corresponding time series as t i , i=1,2,...,n+1; According to the TLE file of the remote sensing satellite, the SGP4 model is used to predict the satellite orbit, and the satellite position and velocity information is quickly obtained through Lagrange interpolation;

[0052] c. Calculate each discrete time point t i Are the satellite and the ground point visible? If they are within the visible range, the satellite is considered to be at the discrete time point t i Visible to the target point;

[0053] d、t i It can be seen that t i-1 When it is not visible, the visible window starts at t i-1 to i Between, when t i-1 It can be seen that t i When it is not visible, the visible window ends at t i-1 to i between.

[0054] Through rough calculation, the approximate interval between the start and end time of the visible time window has been calculated. Using binary search, the precise visible time window can be efficiently calculated. The specific calculation steps are as follows:

[0055] e. If the visible window starts at the time interval (t 0 , t n ), calculate the intermediate time t m The visibility of the satellite to the ground point; if t m It can be seen that its value is assigned to t n If t m If it is not visible, assign its value to t 0 ;

[0056] f. Repeat step e until t n -t 0 <ε, ε is the given error limit;

[0057] g. Get the visible start time t, that is

[0058] The visible start time is thus obtained, and the calculation idea of ​​the visible end time is the same as that of the visible start time.

[0059] S3: Constructing terrain occlusion analysis model

[0060] When a ground target point is located in the coverage area of ​​a remote sensing satellite at a certain moment, the visibility analysis method based on elevation-azimuth is used to analyze the impact of the surrounding mountainous terrain on the coverage area. If the terrain blocks the line of sight of the remote sensing satellite, it means that the ground target point is not in the coverage area of ​​the remote sensing satellite at that moment.

[0061] The terrain elevation data under different environments are obtained through the DEM that can reflect the terrain elevation, and it is read into the simulation software as the data basis for subsequent analysis; the specific steps are:

[0062] (1) Set up equally spaced sampling points within 1000 meters in the direction of line of sight projection;

[0063] (2) Determine the longitude and latitude positions of each sampling point by transforming the coordinate system between the station center coordinate system and the earth coordinate system (λ sample , );

[0064] (3) Using the discrete point elevation information provided by the original DEM data of the mountain, the actual terrain elevation H at each sampling point is obtained through the bilinear interpolation fitting method. sample ;

[0065] (4) Compare the actual terrain of the sampling point with the elevation angle E of the ground target point l_sample The elevation angle E between the remote sensing satellite and the ground target point l The size of E l_sample <E l , then the ground target can be observed by the remote sensing satellite, otherwise the ground target cannot be observed by the remote sensing satellite.

[0066] S4: Performance Test Analysis

[0067] The accuracy and speed of the prediction algorithm of the present invention are verified through a set of simulation examples, and different scenarios are distinguished for test evaluation;

[0068] Scenario 1: Without considering the terrain features, the accuracy is compared with the commercial software STK, and the speed is compared with the classic tracking propagation method; for a ground target point in the canyon, without considering the terrain features, the simulation start time is 00:00:00 (UTC) on November 1, 2024, the simulation end time is 00:00:00 (UTC) on November 2, 2024, the simulation time is 1d, the rough search step is 100s, the precise time step is 0.001s, and the time step of the classic tracking propagation algorithm is set to 0.1s. By comparing the simulation results, it can be obtained that the time difference with the STK visible window in 5 visible time periods is less than 1s; the classic tracking propagation method takes 9.8s, and the algorithm proposed in this invention takes 1.2s.

[0069] Scenario 2: Taking terrain features into consideration and simulating for 1 day, the accuracy and speed of the algorithm used in Scenario 1 were compared. By comparing the simulation results, it can be found that the visible window time in the five visible time periods is significantly reduced, with an average reduction of 37.1%; the algorithm takes 1.5 seconds.

[0070] The parts not described in detail in this invention are prior art.

[0071] The embodiments selected herein for the purpose of disclosing the invention are currently considered to be suitable, but it should be understood that the invention is intended to include all changes and modifications of the embodiments that fall within the scope of the concept and invention.

Claims

1. A fast prediction algorithm for optical remote sensing satellite imaging windows in canyon environments, characterized in that: The specific steps are: S1: Construct visibility model between remote sensing satellite and ground target points Define the maximum geocentric angle corresponding to the minimum observation angle of remote sensing satellite earth observation as d σ ; At a certain time t, the latitude and longitude of the ground target point is The latitude and longitude of the satellite subsatellite point are The calculation formula is: The calculation formula of the geocentric angle between the satellite sub-satellite point and the ground target point is: If d t-s ≤d σ , then at time t, the ground target point is within the instantaneous coverage area of ​​the satellite, otherwise, it is not within the coverage area; i is the orbital plane inclination, ω is the orbital perigee angular distance, Ω0 is the right ascension of the ascending node, and f is the true anomaly; S2: Fast calculation using visible window semi-analysis First determine whether the ground target is on the same side of the earth as the satellite, and then determine whether it is within the visible range of the satellite; use a coarse-precise variable step search, use a large step to preliminarily determine the range of the start and end times, and then use a binary search to accurately calculate the start and end times, and then obtain the visible time window of the regional target; S3: Constructing terrain occlusion analysis model When a ground target point is located in the coverage area of ​​a remote sensing satellite at a certain moment, the visibility analysis method based on elevation angle and azimuth angle is used to analyze the impact of the surrounding high mountain terrain on the coverage area. If the terrain blocks the remote sensing satellite's line of sight, it means that the ground target point is not in the coverage area of ​​the remote sensing satellite at that moment. S4: Performance Test Analysis The accuracy and speed of the prediction algorithm of the present invention are verified through simulation examples, and different scenarios are distinguished for test evaluation; Scenario 1: Without considering terrain features, the accuracy is compared with the commercial software STK, and the speed is compared with the classic tracking propagation method; Scenario 2: Compare the accuracy and speed of the algorithm used in Scenario 1 with those in Scenario 1, taking terrain features into account.

2. The optical remote sensing satellite imaging window rapid prediction algorithm in a canyon environment according to claim 1, characterized in that: The specific calculation steps from rough to precise in S2 are: In the rough calculation, the discrete time step is large, the purpose is to quickly determine the approximate time interval of the visible start time and the visible end time. The specific calculation steps are as follows: a. Given the start and end time of the observation mission, divide the simulation period into several sub-intervals. Let M0 be the point where the satellite passes through the orbital ascending node angle equal to a certain set value each time; from the start time to the first time the satellite passes through M0 is the first section, represented by (tS1, tE1), from the first time the satellite passes through M0 to the second time it passes through M0 is the second section, represented by (tS2, tE2), and so on. The last time the satellite passes through M0 to the end of the simulation is the last section; b. Discretize the entire process of the remote sensing satellite orbit into n equal parts, that is, n+1 discrete points, and record the corresponding time series as t i , i=1,2,…,n+1; According to the TLE file of the remote sensing satellite, the SGP4 model is used to predict the satellite orbit, and the satellite position and velocity information is quickly obtained through Lagrange interpolation; c. Calculate each discrete time point t i Are the satellite and the ground point visible? If they are within the visible range, the satellite is considered to be at the discrete time point t i Visible to the target point; d、t i It can be seen that t i-1 When it is not visible, the visible window starts at t i-1 to i Between, when t i-1 It can be seen that t i When it is not visible, the visible window ends at t i-1 to i between; Through rough calculation, the approximate interval between the start and end time of the visible time window has been calculated. Using binary search, the precise visible time window can be efficiently calculated. The specific calculation steps are as follows: e. If the visible window starts at the time interval (t0, t n ), calculate the intermediate time t m The visibility of the satellite to the ground point; if t m It can be seen that its value is assigned to t n If t m If it is not visible, assign its value to t0; f. Repeat step e until t n -t0<ε, ε is the given error limit; g. Get the visible start time t, that is The visible start time is thus obtained, and the calculation idea of ​​the visible end time is the same as that of the visible start time.

3. The optical remote sensing satellite imaging window rapid prediction algorithm in a canyon environment according to claim 1, characterized in that: The elevation-azimuth visibility analysis method in S3 obtains terrain elevation data under different environments through a DEM that can reflect terrain elevation, and reads it into the simulation software as the data basis for subsequent analysis; the specific steps are: (1) Set up equally spaced sampling points within a range of 100-1000 meters in the direction of line of sight projection; (2) Determine the longitude and latitude of each sampling point by converting the coordinates between the station center coordinate system and the geodetic coordinate system. (3) Using the discrete point elevation information provided by the original DEM data, the actual terrain elevation H at each sampling point is obtained through the bilinear interpolation fitting method. sample ; (4) Compare the actual terrain of the sampling point with the elevation angle E of the ground target point l_sample The elevation angle E between the reconnaissance satellite and the ground target point l The size of E l_sample <E l , then the ground target can be observed by the remote sensing satellite, otherwise the ground target cannot be observed by the remote sensing satellite.

Images