betaFeedback

WorksPhysics and Astronomy

Higher-order composition of short- and long-period effects for improving analytical ephemeris computation

Abstract

The construction of an analytic orbit theory that takes into account the main effects of the Geopotential is notably simplified when splitting the removal of periodic effects in several stages. Conversely, this splitting of the analytical solution into several transformations reduces the evaluation efficiency for dense ephemeris output. However, the advantage is twofold when the different parts of the mean-to-osculating transformation are composed into a single transformation. To show that, Brouwer's solution is extended to the second order of the zonal harmonic of the second degree by the sequential elimination of short- and long-period terms. Then, the generating functions of the different transformations are composed into a single one, from which a single mean-to-osculating transformation is derived. The new, unique transformation notably speeds up the evaluation process, commonly improving evaluation efficiency by at least one third with respect to the customary decomposition of the analytical solution into three different parts.

Introduction

Simple analytic orbit prediction programs still find different astrodynamics applications [1, 2, 3, 4, 5, 6, 7, 8, 9]. They commonly rely on perturbation solutions that are made of secular and periodic terms. The former provide the average evolution of the orbit while the latter are needed to convert the secular elements—also called mean elements—into ephemeris. In the implementation of an analytical ephemeris generator one customarily takes the point of view of using programming techniques that minimize both memory requirements and execution time [10]. However, these two aims are mutually exclusive regarding accuracy (understood as the time span for which the errors remain below a given tolerance).

Minimizing memory requirements is obviously achieved when reducing the truncation order, and, therefore, the accuracy of the perturbation solution. On the other hand, reducing memory needs for a given truncation order can be achieved by splitting the periodic terms of the analytical solution into a sequence of simpler corrections. Separation of the periodic corrections into short- and long-period terms is a natural choice with whole dynamical sense [11, 12]. Moreover, it is well known that the preliminary elimination of the parallax transformation [13, 14] makes notably easier the implementation of the short-period elimination [15, 16]. On the contrary, the long-period elimination is tra ditionally achieved with a single set of corrections, which is obtained either with Brouwer’s traditional method [11], in the reverse normalization style [17], or like Alfriend and Coffey’s halfway option [18, 19]. Splitting the elimination of long-period terms into two simpler transformations is also possible, and helps in pruning away non-essential terms of the rotating-perigee regime, in this way effectively isolating the resonant, long-period terms of the dynamics about the critical inclination. However, the interest of this additional simplification is mostly theoretical since it yields negligible savings in memory storage and usually increases the execution time [20].

While decomposing the transformation from mean to osculating elements into different parts clearly simplifies the periodic terms of the solution by notably reducing their size, the splitting procedure has the undesired side effect of slowing the evaluation of dense output ephemeris. This paradox stems from the fact that the eccentricity and inclination remain constant in the secular variables. Because of that, when the different transformations used in the construction of the analytical perturbation theory are combined into a single transformation, the coefficients of the trigonometric polynomials comprising the periodic corrections only need to be evaluated once, which is done jointly with the initialization of the secular terms of the analytical solution [21]. In this way, the repeated evaluation of the perturbation solution in the computation of ephemeris is notably accelerated. On the contrary, if a sequence of different transformations is used, then these coefficients only remain constant in the first transformation of the sequence, although some of them may remain constant also in the second one [10] —the most favorable case being provided by the reverse normalization scheme [17] in which the only terms that need reevaluation are eccentricity polynomials. This need of repeatedly evaluating coefficients made of inclination and eccentricity polynomials may counterbalance the advantages provided by the simpler form of the sequential periodic corrections, in this way clearly penalizing the efficiency of the analytical theory for dense output.

In order to demonstrate these facts, two alternative higher-order extensions of Brouwer’s classical solution [11] have been implemented. More precisely, since the Earth’s zonal harmonic coefficient of the fifth degree is at least one order of magnitude smaller than the zonal harmonic coefficients of lower degrees, we neglected the contribution of the fifth-degree zonal harmonic from Brouwer’s Geopotential model. In addition, our approach takes the proper calibration of the mean semi-major axis into account [22]. In this way, the dominant long-term secular drift of the position errors in the along-track direction—which is typical of perturbation solutions relying on the physical time as the independent variable [23]—is reduced by at least one order of magnitude with respect to traditional implementations of the Geopotential disturbing effect.

Analytic perturbation solutions. General features

Analytical solutions to orbital perturbation problems are commonly approached in a set of three oscillating and three rotating variables. The latter are naturally angles whereas the former may have different nature. A common choice is the traditional set of Keplerian elements given by the semi-major axis aa, eccentricity ee, inclination II, right ascension of the ascending node Ω\Omega, argument of the periapsis ω\omega, and mean anomaly MM. The perturbation solution is obtained through an analytical transformation to mean elements (a′,e′,I′,Ω′,ω′,M′)(a',e',I',\Omega',\omega',M') such that the first three remain constant, namely,

da′dt=de′dt=dI′dt=0,\frac{\mathrm{d}a'}{\mathrm{d}t}=\frac{\mathrm{d}e'}{\mathrm{d}t}=\frac{\mathrm{d}I'}{\mathrm{d}t}=0,

whereas the last three evolve at constant rates

dΩ′dt=nΩ,dω′dt=nω,dM′dt=nM.\frac{\mathrm{d}\Omega'}{\mathrm{d}t}=n_{\Omega},\qquad\frac{\mathrm{d}\omega'}{\mathrm{d}t}=n_{\omega},\qquad\frac{\mathrm{d}M'}{\mathrm{d}t}=n_{M}.

Because the exact transformation from mean to osculating variables does not exist in general, it is approximated with

T:(a,e,I,Ω,ω,M;ϵ)↦(a′,e′,I′,Ω′,ω′,M′)(1)\mathcal{T}:(a,e,I,\Omega,\omega,M;\epsilon)\mapsto(a',e',I',\Omega',\omega',M') \tag*{(1)}

given by a truncated Taylor series in ϵ\epsilon. The small parameter ϵ\epsilon may be a physical quantity, the most desirable case, or a formal parameter—a token that indicates the strength of the disturbing forces relative to the integrable non-perturbed model. With formal parameters, the analytical solution is constrained to a particular dynamical regime, while physical quantities allow for greater generality.

The perturbation approach is not constrained to the use of Keplerian elements. It can be applied to different sets of singular or non-singular variables. In particular, canonical variables assign uniform dimension to the oscillating-type quantities. In that case, (a,e,I)(a,e,I) are customarily replaced by (L,G,H)(L,G,H). L=μaL=\sqrt{\mu a} is the Delaunay action, with μ\mu denoting the gravitational parameter. G=LηG=L\eta is the specific angular momentum, where η=(1−e2)1/2\eta=(1-e^2)^{1/2}. Finally, H=Gcos⁡IH=G\cos I denotes the third component of the angular momentum vector. The set (L,G,H,ℓ,g,h)(L,G,H,\ell,g,h), with ℓ=M\ell=M, g=ωg=\omega, and h=Ωh=\Omega, is known as the Delaunay canonical variables. They are the action-angle variables in which a complete reduction of the Kepler Hamiltonian is achieved [24, 25, 26].

It is worth remarking that, in the analytical solution by canonical methods, the transformation (1) can be derived from a scalar generating function W=W(L,G,H,ℓ,g,h)W=W(L,G,H,\ell,g,h), simplifying the computational process [27].

Hereafter, we shall limit the discussion to perturbed Keplerian motion and Hamiltonian perturbations in Delaunay variables. Then, the secular frequencies can be written in the general form [28]

nM=n~∑i≥0ϵii!Φi(a0′,e0′,I0′)(2)n_M=\tilde{n}\sum_{i\geq0}\frac{\epsilon^i}{i!}\Phi_i(a'_0,e'_0,I'_0) \tag*{(2)}
nω=n~∑i≥1ϵii!Γi(a0′,e0′,I0′)(3)n_\omega=\tilde{n}\sum_{i\geq1}\frac{\epsilon^i}{i!}\Gamma_i(a'_0,e'_0,I'_0) \tag*{(3)}
nΩ=n~∑i≥1ϵii!Ψi(a0′,e0′,I0′)(4)n_\Omega=\tilde{n}\sum_{i\geq1}\frac{\epsilon^i}{i!}\Psi_i(a'_0,e'_0,I'_0) \tag*{(4)}

where Φi\Phi_i, Γi\Gamma_i, and Ψi\Psi_i, are functions of the initial conditions in prime (mean) variables. Namely, cos⁡I0′=H0′/G0′\cos I'_0=H'_0/G'_0, e0′=(1−G0′2/L0′2)1/2e'_0=(1-G_0^{\prime2}/L_0^{\prime2})^{1/2}, a0′=L0′2/μa'_0=L_0^{\prime2}/\mu, and \tilde{n}=(\mu/a'_0^3)^{1/2}=\mu^2/L_0^{\prime3}. The periodic corrections take the form of truncated multivariate Fourier series in the angle variables, with coefficients that are truncated series in the action variables. Commonly, these are expressed as eccentricity polynomials, with the coefficients given by inclination polynomials [16, 29, 30].

The computation of the constants of the perturbation theory n~\tilde{n}, a0′a'_0, e0′e'_0, I0′I'_0, Ω0′\Omega'_0, ω0′\omega'_0, and M0′M'_0, in mean elements, can be derived from a fit to observations [31], which are obtained either from real data or synthetically generated by a preliminary numerical integration of one or two orbits [22]. Alternatively, the initialization constants can be obtained from an initial state vector in osculating elements by inverting the mean-to-osculating transformation of the perturbation theory [32, 33], an operation that is sometimes replaced by root-finding procedures [34, 35, 36]. Moreover, modern perturbation methods allow for the computation of both the direct and inverse transformations in explicit form [37, 38].

An intrinsic characteristic of perturbation solutions is that, due to the truncation of the series comprised in the solution, they always introduce an error in the secular frequencies. This happens even when the truncation is made to machine precision [17]. In consequence, the errors always undergo a secular drift which, eventually, prevails over the inaccuracies due to the truncation of the periodic corrections. Therefore, it is common practice to compute the periodic terms to one order less than their secular counterparts [11, 31, 25, 16, 10]. However, the proper propagation of the secular frequencies up to a given order requires the initialization of the constants of the perturbation theory to the same truncation order as the secular terms.

In the case of perturbed Keplerian motion, this accuracy can be relaxed in the initialization of nωn_\omega and nΩn_\Omega because they are proportional to ϵ\epsilon, as shown by the lower limit 11 of the summation index in Eqs. (3) and (4). Nevertheless, the higher accuracy is mandatory in the initialization of the secular mean motion nMn_M, for which the summation index in Eq. (2) starts from zero. Neglecting this consideration causes errors in the in-track direction that are inconsistent with the truncation order of the secular terms of the analytical solution [39]. The remedy is to extend the computation of the periodic corrections of the semi-major axis to the same order as the secular terms [40, 41]. When the perturbation solution is computed stepwise, in Brouwer’s seminal style of removing the short-period terms before the long-period ones, these additional computations are limited to the short-period corrections of the semi-major axis.

For Hamiltonian perturbations, rather than computing additional terms of the periodic corrections to the semi-major axis, the semi-major axis can be calibrated to higher-order effects using the energy equation in the clever way proposed by Breakwell and Vagners [22]. The procedure relies on the fact that the energy value for given initial conditions does not change by a transformation of variables. When using orbital elements, the energy equation H≡T+V=E\mathcal{H}\equiv T+V=\mathcal{E}, where TT and VV denote the kinetic and potential energy, respectively, can be written in osculating variables like

H≡−μ2a+ϵP(a,e,I,Ω,ω,M)=E.(5)\mathcal{H}\equiv-\frac{\mu}{2a}+\epsilon\mathcal{P}(a,e,I,\Omega,\omega,M)=\mathcal{E}. \tag*{(5)}

On the other hand, after the complete Hamiltonian reduction, the energy equation takes the form

E=−μ2a′+∑m=1kϵmm!Hm(a′,e′,I′)+O(ϵk+1),(6)\mathcal{E}=-\frac{\mu}{2a'}+\sum_{m=1}^{k}\frac{\epsilon^{m}}{m!}\mathcal{H}_{m}(a',e',I')+\mathcal{O}(\epsilon^{k+1}), \tag*{(6)}

where Hm\mathcal{H}_{m} are the computed Hamiltonian terms. Then, for a given initial state (a0,e0,I0,Ω0,ω0,M0)(a_{0},e_{0},I_{0},\Omega_{0},\omega_{0},M_{0}), we compute the energy value H(a0,e0,I0,Ω0,ω0,M0)=E0\mathcal{H}(a_{0},e_{0},I_{0},\Omega_{0},\omega_{0},M_{0})=\mathcal{E}_{0} exactly from Eq. (5) and replace E=E0\mathcal{E}=\mathcal{E}_{0} in Eq. (6), from which

E0+μ2a0′−∑m=1kϵmm!Hm(a0′,e0′,I0′)=Δ,(7)\mathcal{E}_{0}+\frac{\mu}{2a'_{0}}-\sum_{m=1}^{k}\frac{\epsilon^{m}}{m!}\mathcal{H}_{m}(a'_{0},e'_{0},I'_{0})=\Delta, \tag*{(7)}

where a0′a'_{0}, e0′e'_{0}, and I0′I'_{0} are obtained from the transformation from mean (prime) to osculating variables. If this transformation is known to O(ϵk)\mathcal{O}(\epsilon^{k}), the energy equation (7) is certainly accurate to Δ=O(ϵk+1)\Delta=\mathcal{O}(\epsilon^{k+1}). However, if the transformation is only known to O(ϵk−1)\mathcal{O}(\epsilon^{k-1}), as it is commonly the case, then the error in the energy equation will be only Δ=O(ϵk)\Delta=\mathcal{O}(\epsilon^{k}) due to the propagation of errors in the Keplerian term. The issue is easily fixed by replacing the value a0′a'_{0} obtained from the O(ϵk−1)\mathcal{O}(\epsilon^{k-1}) mean to osculating transformation by the calibrated value

a^0=12μ[−E0+∑m=1kϵmm!Hm(a0′,e0′,I0′)]−1,\hat{a}_{0}=\frac{1}{2}\mu\left[-\mathcal{E}_{0}+\sum_{m=1}^{k}\frac{\epsilon^{m}}{m!}\mathcal{H}_{m}(a'_{0},e'_{0},I'_{0})\right]^{-1},

References

  1. [1]G. Campiti, G. Brunetti, V. Braun, E. Di Sciascio, C. Ciminelli, Orbital kinematics of conjuncting objects in low-earth orbit and opportunities for autonomous observations, Acta Astronautica 208 (2023) 355–366. doi:10.1016/j.actaastro.2023.04.032.DOI
  2. [2]M. Lara, On perturbation solutions in the restricted three-body problem dynamics, Acta Astronautica 195 (2022) 596–604. doi:10.1016/j.actaastro.2022.01.022.DOI
  3. [3]M. Lara, Design of distant retrograde orbits based on a higher order analytical solution, Acta Astronautica 161 (2019) 562–578. doi:10.1016/j.actaastro.2019.01.039.DOI
  4. [4]B. Li, Y. Zhang, J. Huang, J. Sang, Improved orbit predictions using two-line elements through error pattern mining and transferring, Acta Astronautica 188 (2021) 405–415. doi:10.1016/j.actaastro.2021.08.002.DOI
  5. [5]C. T. W. Marshall, Improved orbit predictions using two-line elements, Advances in Space Research 47 (7) (2011) 1107–1115. arXiv:1002.2277, doi:10.1016/j.asr.2010.10.017.
  6. [6]J. F. San-Juan, I. Perez, M. San-martin, E. Vergard, Hybrid SGP4 orbit propagator, Acta Astronautica 137 (2017) 254–260. doi:10.1016/j.actaastro.2017.04.015.DOI
  7. [7]D. A. Vallado, P. Crawford, R. Hujsak, T. S. Kelso, Revisiting Spacetrack Report #3 (AIAA 2006-6753), in: AIAA/AAS Astrodynamics Specialist Conference and Exhibit, Guidance, Navigation, and Control and Co-located Conferences, American Institute of Aeronautics and Astronautics, USA, 2006, pp. 1–88. doi:10.2514/6.2006-6753.DOI
  8. [8]F. R. Hoots, P. W. Schumacher, Jr., R. A. Glover, History of Analytical Orbit Modeling in the U. S. Space Surveillance System, Journal of Guidance, Control, and Dynamics 27 (5) (2004) 174–185. doi:10.2514/1.9161.DOI
  9. [9]F. R. Hoots, R. L. Roehrich, Models for Propagation of the NORAD Element Sets, Project SPACETRACK, Rept. 3, U.S. Air Force Aerospace Defense Command, Colorado Springs, CO (Dec. 1980).DOI
  10. [10]S. Coffey, K. T. Alfriend, An analytical orbit prediction program generator, Journal of Guidance, Control and Dynamics 7 (5) (1984) 575–581. doi:10.2514/3.19897.
  11. [11]D. Brower, Solution of the problem of artificial satellite theory without drag, The Astronomical Journal 64 (1959) 378–397. doi:10.1086/107958.
  12. [12]Y. Kozai, The motion of a close earth satellite, The Astronomical Journal 64 (1959) 367–377. doi:10.1086/107957.DOI
  13. [13]A. Deprit, The elimination of the parallax in satellite theory, Celestial Mechanics 24 (2) (1981) 111–153. doi:10.1007/BF01229192.DOI
  14. [14][14*] M. Lara, J. F. San-Juan, L. M. López-Ochoa, Proper Averaging Via Parallax Elimination (AAS 13-722), in: S. B. Broschart, J. D. Turner, K. C. Howell, F. R. Hoots (Eds.), Astrodynamics 2013, Vol. 150 of Advances in the Astronautical Sciences, American Astronautical Society, Univelit, Inc., P.O. Box 28130, San Diego, California 92198, USA, 2014, pp. 315–331.
  15. [15]A. Deprit, The Main Problem in the Theory of Artificial Satellites to Order Four, Journal of Guidance Control Dynamics 4 (2) (1981) 201–206. doi:10.2514/3.56072.DOI
  16. [16]S. L. Coffey, A. Deprit, Third-Order Solution to the Main Problem in Satellite Theory, Journal of Guidance, Control and Dynamics 5 (4) (1982) 366–371.DOI
  17. [17]M. Lara, Solution to the main problem of the artificial satellite by reverse normalization, Nonlinear Dynamics 101 (2) (2020) 1501–1524. doi:10.1007/s11071-020-05857-3.DOI
  18. [18]K. T. Alfriend, S. L. Coffey, Elimination of the perigee in the satellite problem, Celestial Mechanics 32 (2) (1984) 163–172. doi:10.1007/BF01231123.DOI
  19. [19]M. Lara, J. F. San-Juan, L. M. López-Ochoa, Delaunay variables approach to the elimination of the perigee in Artificial Satellite Theory, Celestial Mechanics and Dynamical Astronomy 120 (1) (2014) 39–56. doi:10.1007/s10569-014-9559-2.DOI
  20. [20]M. Lara, Improving efficiency of analytic orbit propagation (IAC-21-C1.7.2), in: Proceedings of the 72nd International Astronautical Congress (IAC), Dubai, United Arab Emirates, 25–29 October 2021, International Astronautical Federation (IAF), International Astronautical Federation (IAF), 2021, pp. 1–10.
  21. [21]M. Lara, Brouwer’s satellite solution redux, Celestial Mechanics and Dynamical Astronomy 133 (47) (2021) 1–20. doi:10.1007/s10569-021-10043-7.DOI
  22. [22]J. V. Breakwell, J. Vagners, On Error Bounds and Initialization in Satellite Orbit Theories, Celestial Mechanics 2 (1970) 253–264. doi:10.1007/BF01229499.DOI
  23. [23]M. Lara, Note on the analytical integration of circumterrestrial orbits, Advances in Space Research 69 (2022) 4169–4178. doi:10.1016/j.asr.2022.04.007.DOI
  24. [24]C. E. Delaunay, La Théorie du Mouvement de la Lune, Premier volume, Vol. 28 of Mémoires de l’Academie des Sciences de l’Institut Impérial de France., Mallet-Bachelier, Paris, 1860.
  25. [25]A. Deprit, A. Rom, The Main Problem of Artificial Satellite Theory for Small and Moderate Eccentricities, Celestial Mechanics 2 (2) (1970) 166–206.DOI
  26. [26]M. Lara, Analytical and Semianalytical Propagation of Space Orbits: The Role of Polar-Nodal Variables, in: G. Gómez, J. J. Masdemont (Eds.), Astrodynamics Network AstroNet-II: The Final Conference, Vol. 44 of Astrophysics and Space Science Proceedings, Springer, Cham, 2016, pp. 151–166. doi:10.1007/978-3-319-23986-6\_11.DOI
  27. [27]H. Poincaré, Les méthodes nouvelles de la mécanique céleste. Tome 2, Gauthier-Villars et fils (Paris), 1893.
  28. [28]M. Lara, Review of analytical solutions for low earth orbit propagation and study of the precision improvement in the conversion of osculating to mean elements, Technical Report CM 2019/SER0023, Universidad de La Rioja, Logroño, La Rioja (2019).
  29. [29]S. L. Coffey, A. Deprit, Frozen orbits for satellites close to an earth-like planet, Celestial Mechanics and Dynamical Astronomy 59 (1) (1994) 37–72. doi:10.1007/BF00691970.DOI
  30. [30]M. Lara, J. F. San-Juan, D. Hautesserres, HEOSAT: a mean elements orbit propagator program for highly elliptical orbits, CEAS Space Journal 10 (2018) 3–23. doi:10.1007/s12567-017-0152-x.DOI
  31. [31]Y. Kozai, Second-order solution of artificial satellite theory without air drag, The Astronomical Journal 67 (1962) 446–461. doi:10.1086/108753.DOI
  32. [32]B. J. Cain, Determination of mean elements for Brouwer’s satellite theory, Astronomical Journal 67 (1962) 391–392. doi:10.1086/108745.DOI
  33. [33]B. A. Ustinov, Motion of Satellites in Small-Eccentricity Orbits in the Noncentral Gravitational Field of the Earth, Cosmic Research 5 (1967) 159–168.
  34. [34]H. G. Walter, Conversion of osculating orbital elements into mean elements, Astronomical Journal 72 (1967) 994–997. doi:10.1086/110374.DOI
  35. [35]M. Lara, J. F. San-Juan, L. M. López-Ochoa, Averaging Tesseral Effects: Closed Form Relegation versus Expansions of Elliptic Motion, Mathematical Problems in Engineering 2013 (Article ID 570127) (2013) 1–11.DOI
  36. [36]M. Lara, J. F. San-Juan, L. M. López-Ochoa, Efficient semi-analytic integration of GNSS orbits under tesseral effects, Acta Astronautica 102 (2014) 355–366. doi:10.1016/j.actaastro.2013.11.006.DOI
  37. [37]G.-I. Hori, Theory of General Perturbation with Unspecified Canonical Variables, Publications of the Astronomical Society of Japan 18 (4) (1966) 287–296.DOI
  38. [38]A. Deprit, Canonical transformations depending on a small parameter, Celestial Mechanics 1 (1) (1969) 12–30. doi:10.1007/BF01230629.DOI
  39. [39]N. L. Bonavito, S. Watson, H. Walden, An Accuracy and Speed Comparison of the Vinti and Brouwer Orbit Prediction Methods, Technical Report NASA TN D-5203, Goddard Space Flight Center, Greenbelt, Maryland (1969).
  40. [40]R. H. Lyd­dane, C. J. Cohen, Numerical comparison between Brouwer’s theory and solution by Cowell’s method for the orbit of an artificial satellite, Astronomical Journal 67 (1962) 176–177. doi:10.1086/108689.DOI
  41. [41]D. Hautsserres, M. Lara, A fast and efficient algorithm for onboard LEO intermediary propagation, in: Proceedings of the 6th International Conference on Astrodynamics Tools and Techniques, ICATT, ESA, 2016.
  42. [42]J. Henrard, On a perturbation theory using Lie transforms, Celestial Mechanics 3 (1970) 107–120. doi:10.1007/BF01230436.DOI
  43. [43]G. Scheifele, O. Graf, Analytical satellite theories based on a new set of canonical elements, in: Mechanics and Control of Flight Conference, American Institute of Aeronautics and Astronautics, Reston, Virginia, 1974, pp. 1–20. doi:10.2514/6.1974-838.DOI
  44. [44]W. M. Kaula, Analysis of Gravitational and Geometric Aspects of Geodetic Utilization of Satellites, Geophysical Journal 5 (1961) 104–133. doi:10.1111/j.1365-246X.1961.tb00417.x.DOI
  45. [45]W. M. Kaula, Theory of satellite geodesy. Applications of satellites to geodesy, Blaisdell, Waltham, Massachusetts, 1966, reprint: Dover, Mineola, New York, 2000.
  46. [46]M. Lara, Exploring Sensitivity of Orbital Dynamics with Respect to Model Truncation: The Frozen Orbits Approach, in: M. Vasile, E. Minisci, L. Summerer, P. McGinty (Eds.), Stardust Final Conference, Vol. 52 of Astrophysics and Space Science Proceedings, Springer International Publishing, Cham, 2018, pp. 69–83. doi:10.1007/978-3-319-69956-1_4.DOI
  47. [47]M. Lara, R. López, I. Pérez, J. San-Juan, Exploring the long-term dynamics of perturbed Keplerian motion in high degree potential fields, Communications in Nonlinear Science and Numerical Simulations 82 (2020) 105053. doi:10.1016/j.cnsns.2019.105053.
  48. [48]A. Deprit, Lunar theory normalisations, Celestial Mechanics 26 (1982) 9–21. doi:10.1007/BF01233178.
  49. [49]S. L. Coffey, A. Deprit, B. R. Miller, The critical inclination in artificial satellite theory, Celestial Mechanics 39 (4) (1986) 365–406. doi:10.1007/BF01230483.DOI
  50. [50]I. G. Izsak, A note on perturbation theory, The Astronomical Journal 68 (8) (1963) 559–561. doi:10.1086/108180.
  51. [51]K. Aksnes, On the Use of the Hill Variables in Artificial Satellite Theory, Astronomy and Astrophysics 17 (1) (1972) 70–75.
  52. [52]M. Lara, Efficient Formulation of the Periodic Corrections in Brouwer’s Gravity Solution, Mathematical Problems in Engineering vol. 2015 (Article ID 980652) (2015) 1–9. doi:10.1155/2015/980652.DOI
  53. [53]M. Lara, LEO intermediary propagation as a feasible alternative to Brouwer’s gravity solution, Advances in Space Research 56 (3) (2015) 367–376. doi:10.1016/j.asr.2014.12.023.DOI

Paper details

vlnc:W_vv8q2899
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 5

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents