Advertisement

Error Analysis of Sequential Monte Carlo Methods for Transport Problems

  • Rong Kong
  • Jerome Spanier
Conference paper

Abstract

In 1962, Halton introduced a sequential correlated sampling algorithm for the efficient solution of certain matrix problems. We have extended Halton’s method to the solution of certain simple transport problems and the resulting algorithm is capable of producing geometric convergence for these problems. In our algorithm, random walks are processed in groups, called stages, and the result of each stage is a small correction that is added to the solution at the previous stage. It is then of interest to determine conditions that guarantee strict error reduction at each stage forjrarious transport problems. Specifically, if Φ(x) is the true transport solution and \( \tilde \Phi ^{n - 1} (x) \) and \( \tilde \Phi ^n (x) \) are the estimated solutions from the (n-l)st and nth stages, respectively, we demonstrate the existence of a number λ, 0 < λ < 1, which is independent of the stage number n, such that
$$\left\| {{{\tilde \Phi}^n}(x)-\Phi (x)} \right\| \leqslant \lambda \left\| {{{\tilde \Phi}^{n-1}}(x)-\tilde \Phi (x)} \right\| + \in$$
in a certain probabilistic sense, where e is an error term that tends to zero as both the number of terms representing the global solution and the number of random walks per stage tend to infinity. We will indicate how to find such a λ, which is defined in terms of the number of random walks per stage and the coefficients of the transport problem in a rather natural way.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Kong, R., Spanier, J.: Sequential Correlated Sampling Methods for Some Transport Problems. this volumeGoogle Scholar
  2. 2.
    Macrobert, T. M.:Spherical Harmonics. Pergamon Press 1967zbMATHGoogle Scholar
  3. 3.
    Nagy, B. S.:Introduction to Real Functions and Orthogonal Expansions. Oxford University Press 1965zbMATHGoogle Scholar
  4. 4.
    Halton, J.: Sequential Monte Carlo.Proc. Camb. Phil. Soc. 58(1962) 57–78CrossRefzbMATHMathSciNetGoogle Scholar
  5. 5.
    Spanier, J.: Monte Carlo Methods for Flux Expansion Solutions of Transport Problems. to appear inNuclear Science and Engineering Google Scholar
  6. 6.
    Bain, L. J., Engelhardt, M.:Introduction to Probability and Mathematical Statistics. Duxbury Press 1992zbMATHGoogle Scholar
  7. 7.
    Foliand, G. B.:Real Analysis. Wiley-Interscience 1984Google Scholar
  8. 8.
    Ross, S. M.:Introduction to Probability Models. Academic Press 1993zbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Rong Kong
    • 1
  • Jerome Spanier
    • 1
  1. 1.Department of MathematicsClaremont Graduate UniversityClaremontUSA

Personalised recommendations