Satellite orbit semi-major axis flat root long period term modeling method

By performing orbital recursive and sliding filtering on satellites, combined with segmented Fourier fitting, a mathematical model of the semi-major axis flat root long period term of satellite orbit was constructed, and the calculation of semi-major axis flat root long period term of low-orbit satellite orbit was solved. The difficult problem improves the track calculation accuracy and thrust calibration accuracy.

CN118940460APending Publication Date: 2024-11-12CHANGGUANG SATELLITE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410867998.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-01
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The prior art is difficult to accurately calculate the long-term term of the semi-major axis flat root of the low-orbit satellite orbit, resulting in insufficient satellite orbit control accuracy, increasing measurement and control resources, propulsion fuel and time costs.

Method used

By performing long-term orbital recursion on satellites, the transient root of the semi-major axis of the orbit is flat root, and sliding filtering and segmented Fourier fitting methods are used to construct a mathematical model of the flat root of the orbital semi-major axis of the orbital is flat root of the long period term.

Benefits of technology

High-precision modeling of the flat root and long period terms of the satellite orbit semi-major axis is realized, simplifying the calculation process, reducing errors, and improving orbital operation accuracy and thrust calibration accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940460A_ABST
    Figure CN118940460A_ABST
Patent Text Reader

Abstract

The invention relates to a satellite orbit semi-major axis flat root long-period term modeling method, which comprises the following steps: carrying out long-time orbit recursion on a satellite, converting an orbit semi-major axis instantaneous root into an orbit semi-major axis flat root, carrying out sliding filtering processing on the orbit semi-major axis flat root by adopting a sliding filtering method, and carrying out long-period term modeling on the orbit semi-major axis flat root from the perspective of numerical calculation. The optimal calculation time period of the long-period term is obtained by constructing an optimal correlation coefficient calculation method, and furthermore, in the determined calculation time period range, segmented Fourier fitting is performed by taking the geographic longitude of the ascending node of the current circle of orbit as an independent variable and taking the orbit semi-major axis flat root long-period term as a dependent variable, so that the optimal calculation time period of the long-period term is obtained. And finally, a high-precision mathematical model of the orbit semi-major axis flat root long-period term is obtained. According to the method, the calculation process is simple and efficient, the obtained model is high in precision, and accurate modeling of the long-period term of the orbit semi-major axis flat root can be achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of satellite orbit control, and in particular to a method for modeling a satellite orbit semi-major axis flat root long-period term. Background Art

[0002] With the rapid development and construction of low-orbit satellites, the calculation and maintenance requirements for the orbital configuration of the constellation are becoming more and more refined. As the most commonly used orbital variable for orbital control calculations of low-orbit satellites, the calculation of the orbital semi-major axis root often only considers the short-period term, and the long-period term is not fully considered. The impact is that in the process of calculating the orbital semi-major axis root, some long-period term errors are inevitably introduced. The errors caused by this part of the calculation will lead to problems such as insufficient satellite orbit control execution accuracy and inaccurate thrust calibration, resulting in an increase in the cost of measurement and control resources, propulsion fuel, time and manpower consumed in maintaining the constellation.

[0003] For low-orbit satellites, the calculation of the long-period term of the orbital semi-major axis flat root is very difficult due to the extremely complex perturbation forces, and there is currently no effective calculation formula for the calculation. Commonly used calculation methods can generally only consider the impact of low-order perturbations on long periods, and most of the other effects have not been eliminated. Some theories have proposed higher-order calculation formulas, but the calculation formulas of such methods are very complex and the amount of calculation is huge. At the same time, the residual error of the long-period term after calculation is still large, and it still cannot meet many high-precision orbit calculation scenarios. Therefore, there is currently a lack of a simple and efficient calculation method to accurately model the long-period term of the orbital semi-major axis flat root. Summary of the invention

[0004] In order to overcome the shortcomings and difficulties of the above-mentioned prior art, the main purpose of the present invention is to provide a method for efficiently and accurately modeling the long-period term of the semi-major axis flat root of the orbit. Different from the idea of ​​mathematical modeling and analysis based on the perturbation force exerted on the satellite, the present invention simulates the satellite orbit to deduce the long-term orbit change, and performs numerical modeling based on the simulation results, and finally obtains the mathematical expression of the long-period term change of the semi-major axis flat root at different orbital altitudes.

[0005] The technical solution adopted by the present invention is as follows:

[0006] A satellite orbit semi-major axis flat root long period term modeling method, the method comprising the following steps:

[0007] Step 1: Perform long-term orbit recursion based on the initial orbit of the satellite, convert the orbit semi-major axis instantaneous root into the orbit semi-major axis flat root, and use the sliding filter method to perform sliding filtering on the orbit semi-major axis flat root to obtain the orbit semi-major axis flat root long-period term;

[0008] Step 2: perform correlation calculation on the orbital semi-major axis flat root long period term obtained in step 1, and take the maximum correlation coefficient as the goal to obtain the calculation time period of the orbital semi-major axis flat root long period term;

[0009] Step 3: Determine the orbital semi-major axis mean root long period terms distributed in each circle during the entire calculation time period, correspond each orbital semi-major axis mean root long period term to the geographical longitude of the ascending node of the corresponding circle, take the geographical longitude of the ascending node as the independent variable, take the corresponding orbital semi-major axis mean root long period term as the dependent variable, use the piecewise Fourier fitting method to fit and solve the fitting coefficients, and obtain the mathematical model of the orbital semi-major axis mean root long period term.

[0010] The beneficial effects of the present invention are:

[0011] The present invention recurses the long-term orbit of the satellite, converts the orbit semi-major axis instantaneous root into the orbit semi-major axis flat root, adopts the sliding filtering method to perform sliding filtering on the orbit semi-major axis flat root, and then from the perspective of numerical calculation, obtains the optimal calculation time period of the long-period term by constructing the optimal correlation coefficient calculation method, and further within the determined calculation time period range, adopts the geographic longitude of the ascending node of the current orbit as the independent variable, and the orbit semi-major axis flat root long-period term after sliding filtering as the dependent variable, and performs Fourier fitting. In order to improve the fitting accuracy, the segmented Fourier fitting method is adopted, and finally obtains a high-precision fitting mathematical expression, that is, a mathematical model of the orbit semi-major axis flat root long-period term. The present invention has a simple and efficient calculation process, and the obtained model has high accuracy, and can realize accurate modeling of the long-period term of the orbit semi-major axis flat root, effectively solving the problem that the long-period term of the orbit semi-major axis flat root is difficult to calculate and cannot meet the high-precision orbit calculation scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flowchart of a method for modeling a satellite orbit semi-major axis flat root long-period term according to an embodiment of the present invention;

[0013] Figure 2 is a flowchart of the piecewise Fourier fitting method;

[0014] Figure 3 This is the variation of the long period term of the semi-major axis of the orbit with the number of cycles after sliding filter processing;

[0015] Figure 4 It is the graph of correlation coefficient changing with the number of rounds;

[0016] Figure 5 This is the graph of the semi-major axis flat root long period term changing with the number of cycles;

[0017] Figure 6 This is the graph of the semi-major axis flat root long period term changing with the geographic longitude of the ascending node;

[0018] Figure 7 This is the result graph of the fitting residual calculation;

[0019] Figure 8 This is a comparison chart of the elimination effect of the semi-major axis flat root long period term calculation. DETAILED DESCRIPTION

[0020] The technical solution of the present invention will be described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0021] Aiming at the calculation difficulty of long-period terms of orbit semi-major axis flat root, the present invention proposes a modeling method for long-period terms of satellite orbit semi-major axis flat root. The method firstly adopts a sliding filtering method to process the semi-major axis flat root to obtain a long-period variation value, further adopts a method of obtaining an optimal solution by global correlation operation to determine an optimal calculation time period, and finally adopts a piecewise Fourier fitting method to model the semi-major axis flat root long-period terms to obtain an optimal fitting mathematical expression, thereby completing the modeling of the satellite orbit semi-major axis flat root long-period terms.

[0022] Specifically, Figure 1 As shown, the long period term modeling method of the orbit semi-major axis flat root proposed in the embodiment of the present invention mainly includes the following steps:

[0023] Step 1 (S100): Obtain the initial orbit of the satellite, perform long-term orbit recursion based on the initial orbit of the satellite, convert the orbit semi-major axis instantaneous root into the orbit semi-major axis flat root, and use the sliding filtering method to further perform sliding filtering on the orbit semi-major axis flat root to eliminate short-period terms;

[0024] Step 2 (S200): Perform global correlation calculation on the semi-major axis flat root long period term obtained after the sliding filter processing in step 1, with the maximum correlation coefficient as the goal, and obtain the calculation time period of the semi-major axis flat root long period term;

[0025] Step three (S300): According to the calculation time period of the semi-major axis flat root long period term obtained in step two, determine the orbital semi-major axis flat root long period term distributed in each circle within the entire calculation time period, correspond each orbital semi-major axis flat root long period term to the geographical longitude of the ascending node of the corresponding circle, and use the geographical longitude of the ascending node as the independent variable and the corresponding orbital semi-major axis flat root long period term as the dependent variable, use the piecewise Fourier fitting method to fit and solve the fitting coefficient, establish a mathematical expression, and finally obtain the mathematical model of the orbital semi-major axis flat root long period term.

[0026] In step 1, firstly, long-term orbit recursion and flat-root conversion are performed according to the initial orbit of the target to be studied. The orbit recursion uses a high-order earth gravitational field model. The flat-root conversion method commonly used is the flat-root conversion method based on Brouwer, and the expression of the flat-root of the semi-major axis can be expressed as:

[0027] σ t (t) = σ(t) - δσ(t)

[0028] Among them, σ(t) is the instantaneous root of the semi-major axis of the orbit, σ t (t) is the mean root of the orbital semi-major axis, and δσ(t) is the short period term of the orbital semi-major axis taking into account the influence of the main perturbation term of the Earth J2.

[0029] After the above transformation, the semi-major axis square root σ t (t) has a much smaller fluctuation amplitude than the instantaneous root of the semi-major axis of the orbit σ(t), but still contains some short-period terms and long-period terms that change with the latitude argument. For the remaining short-period terms related to the latitude argument, since their variation period is consistent with the orbit period, this embodiment proposes to use a sliding filtering method to filter them out. The filtering window used by the sliding filtering method is the intersection period of the current orbit, and its calculation formula is:

[0030]

[0031] Where i represents the i-th circle in the data, lan(i) represents the geographical longitude of the ascending node corresponding to the i-th circle, and the solution is lan(i) = Ω i -λ i , passing through the right ascension Ω of the ascending node of the i-th circle i The Greenwich hour angle λ of the ascending node relative to the 2000.0 vernal equinox i Do worse and get better. represents the average orbital angular velocity corresponding to the i-th circle, is the average semi-major axis corresponding to the ith circle, and μ represents the Earth’s gravitational field coefficient. is the cosine value of the mean orbital inclination, R e is the radius of the Earth.

[0032] The filtering algorithm expression based on this window is:

[0033]

[0034] Among them, a1(i) is the result of the ith circle obtained after filtering, that is, the semi-major axis square root long period term of the orbit corresponding to the ith circle, a i (t) is the RMS value of the semi-major axis corresponding to the tth moment of the ith cycle.

[0035] In step 2, a global correlation operation is performed on the orbital semi-major axis root long period term al(i). Since a single data cannot be used as the basic unit for correlation calculation, 1 day is selected as the time unit, and the number of data corresponding to 1 day is calculated as:

[0036]

[0037] Among them, Nt is the number of related operation unit data calculated, T a The length of time is 1 day, is the mean of the intersection periods.

[0038] Next, the correlation is calculated with the Nt first group of data and the remaining data in turn to obtain the correlation coefficient. The calculation formula is:

[0039]

[0040] Among them, cov(M,N s ) is the calculated correlation coefficient, M j is the first set of data, used as a benchmark, j is the data sequence number, ranging from 1 to Nt, M j The mean value of satisfy N s For other groups of data to be compared, in N s , s is the step length relative to the first set of data, s∈1,2,..ia-Nt, where ia is the maximum value of round i.

[0041] Through the above operation, we can get ia-Nt correlation coefficients: cov(M,N1),cov(M,N2),...cov(M,N ia-Nt ). Select the maximum correlation coefficient value as the solution result, assuming that s = S r , then S r The number of circles is used as the data time period for subsequent modeling, that is, the calculation time period for the long period term of the mean root of the orbital semi-major axis. Solving this value is essentially to use a simple data algorithm to ensure the density of the fitting data, and then to ensure the fitting accuracy.

[0042] It is worth mentioning that due to the complexity of the trajectory, it is difficult to solve S r For very small orbits, we can make an approximation by making a small adjustment to the initial semi-major axis instantaneous root to avoid s r The problem of low subsequent fitting accuracy caused by the value being too small.

[0043] For S r The data within a circle can be further separated to obtain the long period term of the orbital semi-major axis:

[0044]

[0045] in, is the mean of the semi-major axis mean root long period terms al(i) of all orbits, and l(i) is the calculated mean root long period fluctuation value, which has nothing to do with the change of latitude argument and is only determined by the geographic longitude of the ascending node. Therefore, it can be used as the calculated value of the semi-major axis mean root long period change.

[0046] In step 3, the above steps are used to obtain the distribution in S r The semi-major axis square root long period terms l(1),l(2),...l(S r ), and S r The geographical longitudes of the ascending nodes corresponding to each circle are lan(1), lan(2),...lan(S r ). It is found that the semi-major axis mean root long period term l(i) only changes with the change of the geographic longitude lan(i) of the ascending node, so a function mathematical model can be established with the geographic longitude lan(i) of the ascending node as the independent variable and the orbit semi-major axis mean root long period term l(i) as the dependent variable.

[0047] Since the mathematical model of the function is too complex, the numerical solution of fitting modeling using the discrete data points obtained above is used to establish the function model. Since lan(i)∈[-π,π] and the long-period term of the semi-major axis flat root varies complexly, it needs to be represented by the superposition of multiple frequency functions, so the piecewise Fourier fitting method is chosen here to fit and solve the fitting coefficients.

[0048] Step 31: First, establish an 8th-order Fourier fitting mathematical model as shown below:

[0049] fx=a0+a1*cos(x*w)+b1*sin(x*w)+

[0050] a2*cos(2*x*w)+b2*sin(2*x*w)+a3*cos(3*x*w)+b3*sin(3*x*w)+

[0051] a4*cos(4*x*w)+b4*sin(4*x*w)+a5*cos(5*x*w)+b5*sin(5*x*w)+

[0052] a6*cos(6*x*w)+b6*sin(6*x*w)+a7*cos(7*x*w)+b7*sin(7*x*w)+

[0053] a8*cos(8*x*w)+b8*sin(8*x*w)

[0054] Among them, a0, a1, ... a8 and b0, b1, ... b8 are amplitude coefficients, w is the frequency coefficient, x is the independent variable, here is x = lan(i) ∈ [-π, π], fx is the dependent variable, here is the long-period term fx = l(i).

[0055] The study found that due to the complex changes in l(i), directly fitting the entire range [-π,π] will result in a large residual sum of squares (SSE), and the fitting effect will not meet the requirements. Improving the order of Fourier fitting has little effect, but using segmented fitting can effectively solve this problem. The specific fitting calculation process is as follows: Figure 2 shown.

[0056] Step 32: Determine whether the residual sum of squares exceeds the threshold, if so, execute step 33, otherwise execute step 34;

[0057] Step 33: Divide the independent variable into two segments and substitute them into the 8th-order Fourier fitting mathematical model;

[0058] Steps 3 and 4: Then perform least squares fitting on the source data and use Gaussian elimination method to solve the system of equations to obtain the coefficients of the above Fourier series, i.e. the fitting coefficients.

[0059] The above process is used to calculate the Fourier fitting coefficients of each segment in turn and substitute them into the 8th-order Fourier fitting mathematical model to achieve the modeling of the semi-major axis flat root long-period term.

[0060] The technical scheme and technical effects of the present invention are described below with reference to specific examples.

[0061] The method of the present invention is verified by randomly selecting an orbit in the orbits distributed in the Jilin-1 constellation as the initial orbit of the satellite, wherein the initial orbit is shown in Table 1.

[0062] Table 1 Initial orbit instantaneous root parameters (J2000 inertial system)

[0063] project Parameter Value Time (UTC) December 1, 2023 00:00:00 Semi-major axis 6900.200km Eccentricity 0 inclination 97.4787° Ascending node right ascension 69.5957° Argument of perigee 0 Mean anomaly 0

[0064] Perform 61-day recursion and flat root conversion, and perform sliding filtering according to step 1 to obtain the filtering result. After sliding filtering, the change of the flat root long period term of the orbit semi-major axis with the number of cycles is as follows: Figure 3 shown.

[0065] Then calculate the correlation coefficient in step 2. The calculation result is as follows: Figure 4 As shown, Figure 4 The red star in the middle is the location of the maximum correlation coefficient. The maximum correlation coefficient is 0.9998, and the corresponding circle is S r= 288, and the fitting modeling in step 3 is performed based on the data within the 1-288th cycle. Further separation obtains the semi-major axis flat root long period term as follows Figure 5 shown.

[0066] Then we get the corresponding data of the geographic longitude of the ascending node and the long period term of the semi-major axis square root as follows Figure 6 shown.

[0067] Then, the segmented Fourier fitting calculation is performed according to the method in step 3 to obtain 6 segment fitting functions. The corresponding fitting coefficients are shown in Table 2.

[0068] Table 2 Fourier fitting coefficients

[0069]

[0070]

[0071] The calculated fitting residuals are as follows Figure 7 As shown, the residual sum of squares SSE after fitting is 2.21, and the overall fitting residual is small, meeting the fitting accuracy requirements. In order to verify the effect of the function model, the semi-major axis flat root of other periods after sliding filtering is calculated to eliminate the long period term. The comparison results of the elimination and non-elimination calculated using the function model are shown in Figure 8 As shown in the figure, the blue line represents the long-period value, that is, the semi-major axis flat root when the long-period term is not eliminated, and the red line represents the result after processing, that is, the semi-major axis flat root after the long-period term is eliminated by the function model obtained by the present invention. Figure 8 It can be seen that the semi-major axis flat root long period term calculated by the modeling function has a good effect in predicting and eliminating the semi-major axis flat root long period term of other time periods, and the effectiveness of the method of the present invention has been fully verified.

[0072] The method proposed in the present invention has also been applied in the orbit control process of the Jilin-1 Gaofen 03 and Gaofen 06 satellites of Chang Guang Satellite Technology Co., Ltd., and has shown its application effect in thrust calibration and orbit phase control, with obvious improvements in the accuracy of thrust calibration and phase control.

[0073] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0074] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A satellite orbit semi-major axis flat root long period term modeling method, characterized in that: The following steps are involved: Step 1: Perform long-term orbit recursion based on the initial orbit of the satellite, convert the orbit semi-major axis instantaneous root into the orbit semi-major axis flat root, and use the sliding filter method to perform sliding filtering on the orbit semi-major axis flat root to obtain the orbit semi-major axis flat root long-period term; Step 2: Perform global correlation calculation on the orbital semi-major axis flat root long period term obtained in step 1, and take the maximum correlation coefficient as the goal to obtain the calculation time period of the orbital semi-major axis flat root long period term; Step 3: Determine the orbital semi-major axis mean root long period terms distributed in each circle during the entire calculation time period, correspond each orbital semi-major axis mean root long period term to the geographical longitude of the ascending node of the corresponding circle, take the geographical longitude of the ascending node as the independent variable, take the corresponding orbital semi-major axis mean root long period term as the dependent variable, use the piecewise Fourier fitting method to fit and solve the fitting coefficients, and obtain the mathematical model of the orbital semi-major axis mean root long period term.

2. A satellite orbit semi-major axis flat root long period term modeling method according to claim 1, characterized in that: A high-order Earth gravitational field model is used for long-term orbit recursion.

3. A satellite orbit semi-major axis flat root long period term modeling method according to claim 1 or 2, characterized in that: The Brouwer-based flat root conversion method is used to convert the orbital semi-major axis instantaneous root into the orbital semi-major axis flat root.

4. A satellite orbit semi-major axis flat root long period term modeling method according to claim 1 or 2, characterized in that: In the sliding filtering method, the intersection period of the orbit is used as the filtering window, and the calculation formula is: Where i represents the i-th circle; lan(i) represents the geographical longitude of the ascending node corresponding to the i-th circle, lan(i) = Ω i -λ i ,Ω i is the right ascension of the ascending node of the ith cycle, λ i is the Greenwich hour angle of the ascending node of the ith cycle relative to the 2000.0 vernal equinox; represents the average orbital angular velocity corresponding to the i-th circle, is the average semi-major axis corresponding to the ith circle, μ represents the Earth's gravitational field coefficient; is the cosine value of the mean orbital inclination, R e is the radius of the Earth.

5. A satellite orbit semi-major axis flat root long period term modeling method according to claim 1 or 2, characterized in that: The process of calculating the correlation in step 2 includes the following steps: Step 21: Take 1 day as the time unit and calculate the number of data in 1 day according to the mean of the orbital intersection period: Among them, Nt is the number of related operation unit data calculated, T a The length of time is 1 day, is the mean of the intersection period; Step 22: Use the Nt first set of data and the remaining data to calculate the correlation in turn, and obtain the correlation coefficient cov(M,N s ), the calculation formula is: Among them, M j is the first set of data, j is the data sequence number, ranging from 1 to Nt, M j The mean value of satisfy N s For other groups of data to be compared, N s , s is the step length relative to the first set of data, s∈1,2,..ia-Nt, ia is the maximum value of round i.

6. A satellite orbit semi-major axis flat root long period term modeling method according to claim 1 or 2, characterized in that: The method of piecewise Fourier fitting in step 3 includes the following steps: Step 31: Establish an 8th-order Fourier fitting mathematical model; Step 32: Determine whether the residual sum of squares exceeds the threshold, if so, execute step 33, otherwise execute step 34; Step 33: Divide the independent variable into two segments and substitute them into the 8th-order Fourier fitting mathematical model; Steps 3 and 4: Perform least squares fitting on the source data and use Gaussian elimination method to solve the equations to obtain fitting coefficients. Substitute the fitting coefficients into the 8th-order Fourier fitting mathematical model to obtain the mathematical model of the long-period term of the orbital semi-major axis flat root.