On the convergence rate of the unscented transformation

Article

Abstract

Nonlinear state-space models driven by differential equations have been widely used in science. Their statistical inference generally requires computing the mean and covariance matrix of some nonlinear function of the state variables, which can be done in several ways. For example, such computations may be approximately done by Monte Carlo, which is rather computationally expensive. Linear approximation by the first-order Taylor expansion is a fast alternative. However, the approximation error becomes non-negligible with strongly nonlinear functions. Unscented transformation was proposed to overcome these difficulties, but it lacks theoretical justification. In this paper, we derive some theoretical properties of the unscented transformation and contrast it with the method of linear approximation. Particularly, we derive the convergence rate of the unscented transformation.

Keywords

Unscented transformation Nonlinear transformation  Monte Carlo Linear approximation 

References

  1. Ahn, K. W., Chan, K. S. (2011). Approximate conditional least squares estimation of a nonlinear state-space model via unscented Kalman filter. Technical report# 405 Department of Statistics and Actuarial Science, University of Iowa.Google Scholar
  2. Boyce, W. E., DiPrima, R. C. (2004). Elementary differential equations and boundary value problems (8th ed.). New Jersey: Wiley.Google Scholar
  3. Diekmann, O., Heesterbeek, J. A. P. (2000). Mathematical epidemiology of infectious diseases: Model building, analysis and interpretation. West Sussex: WileyGoogle Scholar
  4. Foster, J., Richards, F. B. (1991). The Gibbs phenomenon for piecewise-linear approximation. The American Mathematical Monthly, 98, 47–49.Google Scholar
  5. Holmquist, B. (1988). Moments and cumulants of the multivariate normal distribution. Stochastic Analysis and Applications, 6, 273–278.Google Scholar
  6. Isserlis, L. (1918). On a formula for the product-moment coefficient of any order of a noraml frequency distribution in any number of variables. Biometrika, 12, 134–139.Google Scholar
  7. Julier, S. J., Uhlmann, J. K. (1997). A new extension of the Kalman filter to nonlinear systems. Proceedings of AeroSense: The 11th International symposium on aerospace/defense sensing, simulation and controls. Orlando, Florida, Vol Multi Sensor Fusion, Tracking and Resource Management II.Google Scholar
  8. Julier, S. J., Uhlmann, J. K. (2004). Unscented filtering and nonlinear estimation. IEEE Review, 92, 401–422.Google Scholar
  9. Milstein, G. N., Tretyakov, M. V. (2004). Stochastic numerics for mathematical physics. Berlin, Heidelberg: Springer.Google Scholar
  10. Shim, H. T., Park, C. H. (2005). The survey of Gibbs phenomenon from Fourier series to hybrid sampling series. Journal of Applied Mathematics and Computing, 17, 719–736.Google Scholar
  11. Simon, D. (2006). Optimal state estimation. New Jersey: Wiley.Google Scholar
  12. Triantafyllopoulos, K. (2003). On the central moments of the multidimensional gaussian distribution. The Mathematical Scientist, 28, 125–128.Google Scholar
  13. Wan, E. A., van der Merwe, R. (2000). The unscented Kalman filter for nonlinear estimation. Adaptive systems for signal processing, communications, and control symposium (AS-SPCC) The IEEE, Oct 2000 Lake Louise, Alberta, pp. 153–158.Google Scholar

Copyright information

© The Institute of Statistical Mathematics, Tokyo 2013

Authors and Affiliations

  1. 1.Division of BiostatisticsMedical College of WisconsinMilwaukeeUSA
  2. 2.Department of Statistics and Actuarial ScienceThe University of IowaIowa CityUSA

Personalised recommendations