Structural Engineering and Mechanics

Techno-Press (테크노프레스)
권호 표지
DOI QR코드

DOI QR Code

Automatic decomposition of unstructured meshes employing genetic algorithms for parallel FEM computations

  • Rama Mohan Rao, A. (Structural Engineering Research Centre, CSIR Campus) ;
  • Appa Rao, T.V.S.R. (Structural Engineering Research Centre, CSIR Campus) ;
  • Dattaguru, B. (Department of Aerospace Engineering, Indian Institute of Science)
  • Received : 2001.12.11
  • Accepted : 2002.10.09
  • Published : 2002.12.25
⟨ Previous Next ⟩

Abstract

Parallel execution of computational mechanics codes requires efficient mesh-partitioning techniques. These mesh-partitioning techniques divide the mesh into specified number of submeshes of approximately the same size and at the same time, minimise the interface nodes of the submeshes. This paper describes a new mesh partitioning technique, employing Genetic Algorithms. The proposed algorithm operates on the deduced graph (dual or nodal graph) of the given finite element mesh rather than directly on the mesh itself. The algorithm works by first constructing a coarse graph approximation using an automatic graph coarsening method. The coarse graph is partitioned and the results are interpolated onto the original graph to initialise an optimisation of the graph partition problem. In practice, hierarchy of (usually more than two) graphs are used to obtain the final graph partition. The proposed partitioning algorithm is applied to graphs derived from unstructured finite element meshes describing practical engineering problems and also several example graphs related to finite element meshes given in the literature. The test results indicate that the proposed GA based graph partitioning algorithm generates high quality partitions and are superior to spectral and multilevel graph partitioning algorithms.

Keywords

References

  1. Altman, E.R., Agarwal, V.K. and Rao, G.R. (1993), "A novel methodology using genetic algorithms for the design of caches and cache replacement policy", Proc. Int. Conf. on Genetic Algorithms, 392-399.
  2. Barnard, S.T. and Simon, H.D. (1994), "A fast multilevel implementation of recursive spectral bisection forpartitioning unstructured problems", Concurrency: Practice and Experience, 6, 101-107. https://doi.org/10.1002/cpe.4330060203
  3. Beasley, D., Bull, D.R. and Martin, R.R. (1993), "An overview of genetic algorithms: part 2, research topics",University Computing, 15(4), 170-181.
  4. Bouhmala, N. and Pahud, M. (1998), "A parallel variant of simulated annealing for optimizing mesh partitionson workstations", Advances in Engineering Software, 29(3-6), 481-485. https://doi.org/10.1016/S0965-9978(98)00013-1
  5. De Keyser, J. and Roose, D. (1992), "Grid partitioning by inertial recursive bisection", Technical Report TW174, Department of Computer Science, Belgium.
  6. Farhat, C. (1988), "A fast simple and efficient automatic FEM domain decomposer", Computers and Structures,28(5), 579-602. https://doi.org/10.1016/0045-7949(88)90004-1
  7. Garey, M.R. and Johnson, D.S. (1979), Computers and Intractability: A guide to the theory of NP-Completeness,Freeman, W.H. and Company, N.Y.
  8. Gil, C., Ortega, J., Diaz, A.F. and Monotoya, M.G. (1998), "Annealing based heuristics and genetic algorithmsfor circuit partitioning in parallel test generation", Future Generation Computing Systems, 14(5), 439-451. https://doi.org/10.1016/S0167-739X(98)00045-4
  9. Goldberg, D.E. (1989), Genetic Algorithms in Search, Optimisation and Machine Learning Addison, Wesley, N.J.
  10. Goldberg, D. and Segrest, P. (1987), "Finite Markov chain analysis of genetic algorithms", Proc. the Fifth Int.Conf. on Genetic Algorithms.
  11. Hendrickson, B. and Leland, R. (1995), "A multilevel algorithm for partitioning graphs", Proc. theSupercomputing 95, ACM.
  12. Holland, J. (1975), Adaptation in Natural and Artificial Systems, University of Michigan Press. Ann Arbor. MI.
  13. Jones, M.T. and Plassmann, P.E. (1994), "Computational results for parallel unstructured mesh computations",Computing Systems in Engineering, 5(4-6), 297-309. https://doi.org/10.1016/0956-0521(94)90013-2
  14. Karypis, G. and Vipin Kumar. (1999), "A fast and high quality multilevel scheme for partitioning irregular graphs", SIAM Journal of Scientific Computing, 20(1), 359-392. https://doi.org/10.1137/S1064827595287997
  15. Kernighan, B.W. and Lin, S. (1970), "An efficient heuristic procedure for partitioning graphs", The Bell SystemTechnical Journal, 29(2), 291-307.
  16. Khan, A.I. and Topping, B.H.V. (1998), "Subdomain generation for parallel finite element analysis", ComputingSystems in Engineering, 4(4-6), 96-129.
  17. Mansoor, N. and Fox, G.C. (1991), "A hybrid genetic algorithm for task allocation in multi-computers", Proc.the 4th Int. Conf. on Genetic Algorithms (Ed. Belew, R.K. and Booker, L.B.), Morgan Kaufmann, 466-473.
  18. Mansour, N. and Fox, G.C. (1994), "Allocating data to distributed memory multiprocessors by geneticalgorithms", Concurrency: Practice and Experience, 6(6), 485-504. https://doi.org/10.1002/cpe.4330060602
  19. Miller, G.L., Teng, S., Thurston, W. and Vavasis, S.A. (1998), "Geometric separators for finite element meshes",SIAM Journal of Scientific Computing, 19(2), 364-386. https://doi.org/10.1137/S1064827594262613
  20. Mitchel, W.J., Steadman, J.P. and Liggett, R.S. (1976), "Synthesis and optimisation of small rectangular floorplans", Environment and Planning, 3, 37-70. https://doi.org/10.1068/b030037
  21. Pain, C.C., De Oliveira, C.R.E. and Goddard, A.J.H. (1999), "A neural network graph partitioning procedure forgrid based domain decomposition", Int. J. Numer. Methods Eng., 44, 593-613. https://doi.org/10.1002/(SICI)1097-0207(19990220)44:5<593::AID-NME516>3.0.CO;2-0
  22. Park, K. and Carter, B. (1995), "On the effectiveness of genetic search in combinatorial optimisation", Proc. the10th ACM Symposium on Applied Computing, Genetic Algorithms and Optimisation Track.
  23. Plaskacz, E.J., Ramirez, M.R. and Gupta, S. (1994), "Non-linear explicit transient finite element analysis on theintel delta", Computing Systems in Engineering, 5, 1-17. https://doi.org/10.1016/0956-0521(94)90033-7
  24. Punch, W., Goodman, E., Pei, M., Lai, C.S., Hovland, P. and Enbody, R. (1993), "Intelligent clustering of highdimensionality data using genetic algorithms", Proc. Int. Conf. on Genetic Algorithms, 557-564 .
  25. Rama Mohan Rao, A. (2001), "Efficient parallel processing algorithms for nonlinear dynamic analysis", Ph.D.Thesis Submitted to Indian Institute of Science, Bangalore. India.
  26. Rama Mohan Rao, A., Appa Rao, T.V.S.R. and Dattaguru, B. (1998), "Load balancing for parallel finite elementanalysis employing artificial neural networks", In CDROM Proc. Int. Conf. on Theoretical AppliedComputational and Experimental Mechanics.
  27. Rama Mohan Rao, A., Umesha, P.K. and Loganathan, K. (1998), "PSTRAIN: A graphic user interface forparallel nonlinear dynamic analysis of structures", SERC Technical Report NO. GAP 0641-PPG-TR-98-02.
  28. Simon, H.D. (1991), "Partitioning of unstructured problems for parallel processing", Computing Systems inEngineering, 2, 135-148. https://doi.org/10.1016/0956-0521(91)90014-V
  29. Soper, A.J., Walshaw, C. and Cross, M. (2000), "A combined evolutionary search and multilevel optimisationapproach to graph partitioning", Mathematics Research Report 00/IM/58.
  30. Walshaw, C. and Cross, M. (1999), "Parallel optimisation algorithms for multilevel mesh partitioning",Mathematics Research Report 99/IM/44.
  31. Wendl, S. (1996), "A seed based decomposition algorithm employing genetic algorithms", Report No. EPCCSS96-01, University of Edinburgh, UK.
  32. Williams, R.D. (1991), "Performance of dynamic load balancing algorithms for unstructured mesh calculations",Concurrency, 3, 457-481. https://doi.org/10.1002/cpe.4330030502

Cited by

  1. Distributed evolutionary multi-objective mesh-partitioning algorithm for parallel finite element computations vol.87, pp.23-24, 2009, https://doi.org/10.1016/j.compstruc.2009.05.006
  2. Parallel mesh-partitioning algorithms for generating shape optimised partitions using evolutionary computing vol.40, pp.2, 2009, https://doi.org/10.1016/j.advengsoft.2008.03.017
  3. A new parallel overlapped domain decomposition method for nonlinear dynamic finite element analysis vol.81, pp.26-27, 2003, https://doi.org/10.1016/S0045-7949(03)00312-2
  4. Multilevel graph partitioning: an evolutionary approach vol.56, pp.5, 2005, https://doi.org/10.1057/palgrave.jors.2601837
  5. A mesh partitioning algorithm for generation of shape optimized submeshes using evolutionary computing vol.3, pp.3, 2008, https://doi.org/10.1556/Pollack.3.2008.3.8
  6. An optimized mesh partitioning in FEM based on element search technique vol.23, pp.5, 2019, https://doi.org/10.12989/cac.2019.23.5.311