Implicit-Runge–Kutta-based methods for fast, precise, and scalable uncertainty propagation

Abstract

A large class of methods for propagating the state of an object and its associated uncertainty require the propagation of an ensemble of particles or states through nonlinear dynamics. Existing sigma point- or particle-based methods for uncertainty propagation solve the ensemble of initial value problems one-by-one without exploiting the proximity of the initial conditions. In this paper, we demonstrate how implicit Runge–Kutta methods can be modified to solve initial-value-problem ensembles collectively in order to significantly reduce the computational cost of uncertainty propagation. Examples and variations of this approach are given in the context of the perturbed two-body problem of orbital mechanics that arises in space-object tracking, conjunction analysis, maneuver detection, and other Astrodynamics functions. Particular attention is given to the accuracy of the solution to the initial-value-problem ensemble when the particles are propagated collectively.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2

Notes

  1. 1.

    In the Gaussian case, a proper characterization of uncertainty is sometimes called covariance realism or covariance consistency (Drummond et al. 2006). In the more general case, it is called uncertainty realism or uncertainty consistency.

  2. 2.

    Uncertainty in a model parameter can be treated in many ways, including as part of the estimation process, assuming that the model parameter is observable.

  3. 3.

    The Kalman filter optimally characterizes the state and uncertainty in the linear Gaussian case (Kalman 1960) and the extended Kalman filter approximately characterizes the state in the Gaussian and (mildly) nonlinear case (Jazwinski 1970; Gelb 1974).

  4. 4.

    The solution of the unperturbed Kepler problem, for example, requires solving only a single transcendental equation in the universal anomaly. Efficient methods for doing so are well known (Montenbruck and Gill 2005).

  5. 5.

    Local error is incurred over each time step, and results from a combination of truncation error and round-off error. Accumulation and transport of local error leads to the global error. In the context of many applications including orbital propagation (Shampine 2005; Aristoff et al. 2014b), local error control gives rise to global error control and more efficient propagation compared to methods that employ fixed steps sizes.

  6. 6.

    For super-convergent IRK methods such as Gauss–Legendre IRK, estimation of the local error can be expensive. Hence, the reuse of error estimates leads to a considerable speedup.

  7. 7.

    We have in other work (Horwood et al. 2014b) developed metrics to assess the breakdown of specific PDFs, and applied said metrics to non-linear uncertainty propagation methods used for space surveillance (Horwood et al. 2014a).

  8. 8.

    The equinoctial elements \((a,h,k,p,q,\ell )\) are defined in terms of the Keplerian elements \((a,e,i,\varOmega ,\omega ,M)\) by

    $$\begin{aligned} \begin{aligned}&a = a, \quad h = e \sin (\omega + \varOmega ), \quad k = e \cos (\omega + \varOmega ), \\&p = \tan \tfrac{i}{2} \sin \varOmega , \quad q = \tan \tfrac{i}{2} \cos \varOmega , \quad \ell = M + \omega + \varOmega , \end{aligned} \end{aligned}$$
    (11)

    where \(a\) is the semi-major axis, \(e\) the eccentricity, \(i\) the inclination, \(\varOmega \) the longitude of the ascending node, \(\omega \) the argument of periapsis, and \(M\) the mean anomaly (Broucke and Cefola 1972). Note that unlike the Keplerian elements, the equinoctial elements are non-singular for zero eccentricity and inclination.

  9. 9.

    Each particle is transformed to Cartesian coordinates in which the IVP (1)–(2) is most easily expressed.

  10. 10.

    By neglecting all but the dominant term in the spherical harmonic expansion of the Earth’s gravitational field, the equations of motion permit an analytical solution.

References

  1. Aristoff, J.M.: Method and system for propagating the state of an object and its uncertainty. U.S. Patent Number 8,909,588 B2 (2014)

  2. Aristoff, J.M., Poore, A.B.: Implicit Runge–Kutta methods for orbit propagation. In: Proceedings of the AIAA/AAS Astrodynamics Specialist Conference, Minneapolis, MN, August. Paper AIAA 2012–4880 (2012)

  3. Aristoff, J.M., Horwood, J.T., Poore, A.B.: Implicit Runge–Kutta methods for uncertainty propagation. In: Proceedings of the Advanced Maui Optical and Space Surveillance Technologies Conference, Wailea, HI, September (2012)

  4. Aristoff, J.M., Horwood, J.T., Poore, A.B.: Orbit and uncertainty propagation: a comparison of Gauss–Legendre-, Dormand–Prince-, and Chebyshev–Picard-based approaches. Celest. Mech. Dyn. Astron. 118(1), 13–28 (2014a)

    Article  ADS  Google Scholar 

  5. Aristoff, J.M., Horwood, J.T., Singh, N., Poore, A.B.: Error estimation and control for efficient and reliable orbit (and uncertainty) propagation. In: Proceedings of the 24th AAS/AIAA Space Flight Mechanics Meeting, Santa Fe, NM, January. Paper AIAA 2014–346 (2014b)

  6. Bai, X., Junkins, J.L.: Modified Chebyshev–Picard iteration methods for orbit propagation. J. Astronaut. Sci. 3, 1–27 (2011)

    Google Scholar 

  7. Bar-Shalom, Y., Li, X.-R., Kirubarajan, T.: Estimation with Applications to Tracking and Navigation. Wiley, New York (2001)

    Google Scholar 

  8. Beylkin, G., Sandberg, K.: ODE solvers using bandlimited approximations. J. Comput. Phys. 265, 156–171 (2014)

    Article  ADS  MathSciNet  Google Scholar 

  9. Blackman, S., Popoli, R.: Design and Analysis of Modern Tracking Systems. Artech House, Boston (1999)

    Google Scholar 

  10. Bradley, B.K., Jones, B.A., Beylkin, G., Axelrad, P.: A new numerical integration technique in astrodynamics. In: Proceedings of the 22nd Annual AAS/AIAA Spaceflight Mechanics Meeting, Charleston, SC, 30 Jan– 2 Feb, AAS 12–216 (2012)

  11. Bradley, B.K., Jones, B.A., Beylkin, G., Sandberg, K., Axelrad, P.: Bandlimited implicit Runge–Kutta integration for astrodynamics. Celest. Mech. Dyn. Astron. 119, 143–168 (2014)

    Article  ADS  MATH  MathSciNet  Google Scholar 

  12. Broucke, R.A., Cefola, P.J.: On the equinoctial orbit elements. Celest. Mech. 5, 303–310 (1972)

    Article  ADS  MATH  Google Scholar 

  13. Butcher, J.C.: Numerical Methods for Ordinary Differential Equations. Wiley, West Sussex (2008)

    Google Scholar 

  14. Chan, F.K.: Spacecraft Collision Probability. The Aerospace Press, El Segundo, CA (2008)

    Google Scholar 

  15. DeMars, K.J., Jah, M.K., Cheng, Y., Bishop, R.H.: Methods for splitting Gaussian distributions and applications within the AEGIS filter. In: Proceedings of the 22nd AAS/AIAA Space Flight Mechanics Meeting, Charleston, SC, Feb, Paper AAS-12-261 (2012)

  16. Drummond, O.E., Perrella, Jr. A.J., Waugh, S.: On target track covariance consistency. In: SPIE Proceedings: Signal and Data Processing of Small Targets 2006, vol. 6236 (2006)

  17. Gelb, A.: Applied Optimal Estimation. M.I.T. Press, Cambridge, MA (1974)

    Google Scholar 

  18. Hairer, E., Wanner, G.: Solving ordinary differential equations II: stiff and differential-algebraic problems. In: Springer Series in Computational Mathematics, vol. 14, 2nd edn. Springer, Berlin (2010)

  19. Hairer, E., Norsett, S.P., Wanner, G.: Solving ordinary differential equations I: nonstiff problems. In: Springer Series in Computational Mathematics, vol. 8, 2nd edn. Springer, Berlin (2009)

  20. Hero, A.O., Castañón, D., Cochran, D., Kastella, K.: Foundations and Applications of Sensor Management. Springer, New York (2008)

    Google Scholar 

  21. Horwood, J.T., Poore, A.B.: Adaptive Gaussian sum filters for space surveillance. IEEE Trans. Autom. Control 56(8), 1777–1790 (2011)

    Article  MathSciNet  Google Scholar 

  22. Horwood, J.T., Poore, A.B.: Orbital state uncertainty realism. In: Proceedings of the Advanced Maui Optical and Space Surveillance Technologies Conference, Wailea, HI (2012)

  23. Horwood, J.T., Poore, A.B.: Gauss von Mises distribution for improved uncertainty quantification in space situational awareness. SIAM J. Uncertain. Quantif. 2, 276–304 (2014)

    Article  MATH  MathSciNet  Google Scholar 

  24. Horwood, J.T., Aragon, N.D., Poore, A.B.: Gaussian sum filters for space surveillance: theory and simulations. J. Guid. Control Dyn. 34(6), 1839–1851 (2011)

    Article  ADS  Google Scholar 

  25. Horwood, J.T., Aristoff, J.M., Singh, N., Poore, A.B.: A comparative study of new non-linear uncertainty propagation methods for space surveillance. In: Drummond, O.E. (ed.) SPIE Proceedings: Signal and Data Processing of Small Targets 2014, vol 9092 (2014a)

  26. Horwood, J.T., Aristoff, J.M., Singh, N., Poore, A.B., Hejduk, M.D.: Beyond covariance realism: a new metric for uncertainty realism. In: Drummond, O.E. (ed.) SPIE Proceedings: Signal and Data Processing of Small Targets 2014, vol. 9092 (2014b)

  27. Iserles, A.: A First Course in the Numerical Analysis of Differential Equations. Cambridge University Press, Cambridge (2004)

    Google Scholar 

  28. Ito, K., Xiong, K.: Gaussian filters for nonlinear filtering problems. IEEE Trans. Autom. Control 45(5), 910–927 (2000)

    Article  MATH  MathSciNet  Google Scholar 

  29. Jazwinski, A.H.: Stochastic Processes and Filtering Theory. Dover, New York (1970)

    Google Scholar 

  30. Jones, B.A.: Orbit propagation using Gauss–Legendre collocation. In: Proceedings of the AIAA/AAS Astrodynamics Specialist Conference, Minneapolis, MN, Paper AIAA 2012–4967 (2012)

  31. Jones, B.A., Doostan, A., Born, G.H.: Nonlinear propagation of orbit uncertainty using non-intrusive polynomial chaos. J. Guid. Control Dyn. 36(2), 430–444 (2013). doi:10.2514/1.57599

    Article  ADS  Google Scholar 

  32. Julier, S.J., Uhlmann, J.K.: Method and apparatus for fusing signals with partially known independent error components. U.S. Patent Number 6,829,568 B2 (2004a)

  33. Julier, S.J., Uhlmann, J.K.: Unscented filtering and nonlinear estimation. Proc. IEEE 92, 401–422 (2004b)

    Article  Google Scholar 

  34. Kalman, R.E.: A new approach to linear filtering and prediction problems. J. Basic Eng. 82(1), 35–45 (1960)

    Article  Google Scholar 

  35. Kelley, C.T.: Iterative methods for linear and nonlinear equations. Frontiers in Applied Mathematics, vol. 16. SIAM (1995)

  36. Le Maître, O.P., Knio, O.M.: Spectral Methods for Uncertainty Quantification with Applications to Computational Fluid Dynamics. Springer, London (2010)

    Google Scholar 

  37. Mahler, R.P.S.: Statistical Multisource-Multitarget Information Fusion. Artech House, Boston (2007)

    Google Scholar 

  38. Montenbruck, O., Gill, E.: Satellite Orbits: Models, Methods, and Applications. Springer, Berlin (2005)

    Google Scholar 

  39. Pavlis, N.K., Holmes, S.A., Kenyon, S.C., Factor, J.K.: An Earth gravitational model to degree 2160: EGM2008. In: Proceedings of the European Geosciences Union General Assembly, Vienna, Austria (2008)

  40. Poore, A.B.: Multidimensional assignment formulation of data association problems arising from multitarget tracking and multisensor data fusion. Comput. Optim. Appl. 3, 27–57 (1994)

    Article  MATH  MathSciNet  Google Scholar 

  41. Ristic, B., Arulampalam, S., Gordon, N.: Beyond the Kalman Filter: Particle Filters for Tracking Applications. Artech House, Boston (2004)

    Google Scholar 

  42. Shampine, L.F.: Numerical Solution of Ordinary Differential Equations. Chapman and Hall, New York (1994)

    Google Scholar 

  43. Shampine, L.F.: Error estimation and control for ODEs. J. Sci. Comput. 25, 3–16 (2005)

    Article  MATH  MathSciNet  Google Scholar 

  44. Terejanu, G., Singla, P., Singh, T., Scott, P.D.: Uncertainty propagation for nonlinear dynamic systems using Gaussian mixture models. J. Guid. Control Dyn. 31(6), 1623–1633 (2008)

    Article  ADS  Google Scholar 

  45. United States Air Force Chief Scientist: Report on technology horizons: a vision for Air Force Science & Technology during 2010–2030. Technical report, AF/ST-TR-10-01-PR, vol. 1 (2010)

  46. Xiu, D.: Numerical Methods for Stochastic Computations: A Spectral Method Approach. Princeton University Press, Princeton (2010)

    Google Scholar 

Download references

Acknowledgments

The authors thank Navraj Singh for fruitful discussions and the reviewers for helpful feedback. This work was funded, in part, by a Phase II STTR from the Air Force Office of Scientific Research (FA9550-12-C-0034) and a grant from the Air Force Office of Scientific Research (FA9550-11-1-0248).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Jeffrey M. Aristoff.

Appendix: Convergence of the fixed-point iterations—proof and example

Appendix: Convergence of the fixed-point iterations—proof and example

Fixed-point methods for solving (3) or (6) are sometimes used because they do not require taking derivatives of \(f\) or \(g\), which can be computationally prohibitive. However, unlike Newton or quasi-Newton methods, care must be taken when choosing the step size in order to ensure that the iterations converge on each (time) step.

Convergence of the nonlinear system (3) via fixed-point methods is guaranteed provided \(f\) fulfills the Lipshitz condition

$$\begin{aligned} \Vert f(t,y)-f(t,x)\Vert \le L\Vert y - x\Vert \end{aligned}$$
(12)

and \(s\cdot h \cdot (\max |a_{ij}|) \cdot L < 1\), where \(L\) is the Lipshitz constant for the function \(f\), and where \(s\) is the number of stages in the IRK method (Shampine 1994). Similarly, one may apply the contraction mapping theorem to show that convergence of a second-order nonlinear system is guaranteed provided \(g\) fulfills the Lipshitz condition (12), \(s\cdot h \cdot (\max |a_{ij}|) \cdot L' < 1\) and \(s\cdot h^2 \cdot (\max |\bar{a}_{ij}|) \cdot L < 1\), where \(L\) and \(L'\) are the Lipshitz constants for the function \(g\) (corresponding to the position and velocity components, respectively). Hence, there exists a step size \(h\) below which an IRK method will converge (given a well-behaved \(f\) or \(g\)), regardless of \(L\) and \(L'\). Implementations of IRK can therefore detect if the iterative method used to solve (3) or (6) is converging, and if it is not, then the step size can be reduced.

For the two-body problem, we may estimate the magnitude of the step size \(h\) below which the iterations will converge by considering the case of unperturbed Keplerian dynamics, wherein \(g(t,r) = -(\mu /r^3) r\), \(\mu \) is the gravitational coefficient, and \(r\) is the distance from the orbiting object to the center of the Earth. Consider two nearby orbits, \(r_1\) and \(r_2=r_1+\varDelta r\). By retaining only the first two terms in the Taylor series expansion of \(g(t,r_2)\), the Lipschitz condition (12) becomes

$$\begin{aligned} \Vert g(t,r_2)-g(t,r_1)\Vert \approx \left\| \varDelta r \frac{\partial g}{\partial r} \right\| = L\Vert r_2+\varDelta r-r_1\Vert =L\Vert \varDelta r\Vert . \end{aligned}$$
(13)

Since \(\Vert \frac{\partial g}{\partial r}\Vert \le \mu /(27r^3)\), we find that \(L\le \mu /(27r^3)\). In addition, since \(\partial g / \partial r' = 0\), we have that \(L'=0\). Thus, the Lipshitz constant is small. Convergence of the fixed-point iterations require a step size \(h\) that satisfies

$$\begin{aligned} h < \frac{(3\sqrt{3})r^{3/2}}{\sqrt{\mu \cdot s \cdot \max _{i,j=1,\ldots ,s}{|\bar{a}_{ij}|}}}. \end{aligned}$$
(14)

By making the observation that \(0.73<\sqrt{s\cdot \max _{i,j=1,\ldots , s}|\bar{a}_{ij}|}<1.00\) (for Gauss–Legendre IRK methods having 50 or fewer stages per step), a conservative estimate of \(h_{ max}\) can be made:

$$\begin{aligned} h_{ max} = \frac{(3\sqrt{3})r^{3/2}}{\sqrt{\mu }}. \end{aligned}$$
(15)

For nearly-circular orbits, \(r\approx a\), where \(a\) is the semi-major axis. Since the mean motion and the orbital period \(T\) are related via \(T=2\pi /n\) where \(n=\sqrt{\mu /a^3}\), we can rewrite (15) as

$$\begin{aligned} \frac{h_{ max}}{T} = \frac{3\sqrt{3}}{2\pi }\approx 0.8. \end{aligned}$$
(16)

This result suggests that time steps up to \(4/5\) of an orbital period should converge, which is much greater than the time steps taken in the examples presented in Sect. 3.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Aristoff, J.M., Horwood, J.T. & Poore, A.B. Implicit-Runge–Kutta-based methods for fast, precise, and scalable uncertainty propagation. Celest Mech Dyn Astr 122, 169–182 (2015). https://doi.org/10.1007/s10569-015-9614-7

Download citation

Keywords

  • Implicit Runge–Kutta (IRK)
  • Initial value problem (IVP)
  • Orbit and uncertainty propagation
  • Satellites