Learning Overlap Optimization for Domain Decomposition Methods

  • Steven Burrows
  • Jörg Frochte
  • Michael Völske
  • Ana Belén Martínez Torres
  • Benno Stein
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7818)

Abstract

The finite element method is a numerical simulation technique for solving partial differential equations. Domain decomposition provides a means for parallelizing the expensive simulation with modern computing architecture. Choosing the sub-domains for domain decomposition is a non-trivial task, and in this paper we show how this can be addressed with machine learning. Our method starts with a baseline decomposition, from which we learn tailored sub-domain overlaps from localized neighborhoods. An evaluation of 527 partial differential equations shows that our learned solutions improve the baseline decomposition with high consistency and by a statistically significant margin.

Keywords

domain decomposition numerical simulation 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Bjorstad, P., Hvidsten, A.: Iterative Methods for Substructured Elasticity Problems in Structural Analysis. In: Glowinski, R. (ed.) Proceedings of the First International Symposium on Domain Decomposition Methods for Partial Differential Equations, pp. 301–312. SIAM, Paris (1987)Google Scholar
  2. 2.
    Brady, T.F., Yellig, E.: Simulation Data Mining: A New Form of Computer Simulation Output. In: Kuhl, M.E., Steiger, N.M., Armstrong, F.B., Joines, J.A. (eds.) Proceedings of the Thirty-Seventh Winter Simulation Conference, pp. 285–289. ACM, Orlando (2005)Google Scholar
  3. 3.
    Brenner, S.C., Scott, L.R.: The Mathematical Theory of Finite Element Methods, 2nd edn. Springer, Berlin (2002)CrossRefMATHGoogle Scholar
  4. 4.
    Burrows, S., Stein, B., Frochte, J., Wiesner, D., Müller, K.: Simulation Data Mining for Supporting Bridge Design. In: Christen, P., Li, J., Ong, K.L., Stranieri, A., Vamplew, P. (eds.) Proceedings of the Ninth Australasian Data Mining Conference, pp. 163–170. ACM, Ballarat (2011)Google Scholar
  5. 5.
    Canuto, C.G., Hussaini, M.Y., Quarteroni, A., Zang, T.A.: Spectral Methods: Evolution to Complex Geometries and Applications to Fluid Dynamics, 1st edn. Scientific computation. Springer, Heidelberg (2007)MATHGoogle Scholar
  6. 6.
    Gerold, F., Beucke, K., Seible, F.: Integrative Structural Design. Journal of Computing in Civil Engineering 26(6), 720–726 (2012)CrossRefGoogle Scholar
  7. 7.
    Hall, M., Frank, E., Holmes, G., Pfahringer, B., Reutemann, P., Witten, I.H.: The WEKA Data Mining Software: An Update. SIGKDD Explorations 11(1), 10–18 (2009)CrossRefGoogle Scholar
  8. 8.
    Mei, L., Thole, C.A.: Data Analysis for Parallel Car-Crash Simulation Results and Model Optimization. Simulation Modelling Practice and Theory 16(3), 329–337 (2008)CrossRefGoogle Scholar
  9. 9.
    Press, W.H., Flannery, B.P., Teukolsky, S.A., Vetterling, W.T.: LU Decomposition and its Applications. In: Numerical Recipes in Fortran: The Art of Scientific Computing, 2nd edn., pp. 34–42. Cambridge University Press, Cambridge (1992)Google Scholar
  10. 10.
    Quarteroni, A., Valli, A.: Domain Decomposition Methods for Partial Differential Equations, 1st edn. Numerical Mathematics and Scientific Computation. Oxford Science Publications, New York City (1999)MATHGoogle Scholar
  11. 11.
    Stein, B., Curatolo, D.: Selection of Numerical Methods in Specific Simulation Applications. In: del Pobil, A.P., Mira, J., Ali, M. (eds.) IEA/AIE 1998. LNCS, vol. 1416, pp. 918–927. Springer, Heidelberg (1998)CrossRefGoogle Scholar
  12. 12.
    Toselli, A., Widlund, O.: Domain Decomposition Methods – Algorithms and Theory., 1st edn. Springer Series in Computational Mathematics, vol. 34. Springer, Heidelberg (2004)Google Scholar
  13. 13.
    Versteeg, H.K., Malalasekera, W.: An Introduction to Computational Fluid Dynamics: The Finite Volume Method, 2nd edn. Pearson Education, Essex (2007)Google Scholar
  14. 14.
    Yang, S.-H., Hu, B.-G.: Reformulated Parametric Learning based on Ordinary Differential Equations. In: Huang, D.-S., Li, K., Irwin, G.W. (eds.) ICIC 2006. LNCS (LNAI), vol. 4114, pp. 256–267. Springer, Heidelberg (2006)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Steven Burrows
    • 1
  • Jörg Frochte
    • 2
  • Michael Völske
    • 1
  • Ana Belén Martínez Torres
    • 2
  • Benno Stein
    • 1
  1. 1.Web Technology and Information SystemsBauhaus-Universität WeimarWeimarGermany
  2. 2.Electrical Engineering and Computer ScienceBochum University of Applied ScienceBochumGermany

Personalised recommendations