Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 112
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED, GPU AND CLOUD COMPUTING FOR ENGINEERING Edited by:
Paper 3
Structural reliability computations with the use of the DOProC method and parallel processing J. Koktan, M. Krejsa, J. Brozovsky and P. Janas
VSB – Technical University of Ostrava, Faculty of Civil Engineering, Ostrava, Czech Republic J. Koktan, M. Krejsa, J. Brozovsky, P. Janas, "Structural reliability computations with the use of the
DOProC method and parallel processing", in , (Editors), "Proceedings of the Sixth International Conference on Parallel, Distributed, GPU and Cloud Computing for Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 3, 2019. doi:10.4203/ccp.112.3
Keywords: reliability, DOProC, parallel processing, civil engineering, structural mechanics.
Summary
Analysis of structural reliability often requires the use of Monte Carlo-based simulation
methods. The DOProC (Direct Optimized Probabilistic Calculation) method is an approach
which can be used as an alternative to the Monte Carlo method. It uses the basic principles of
statistics computations and it does not use random numbers. All computations are predictable.
The simplest implementation of the DOProC method requires a large number of repeated
computations which can be of several orders higher than the number of simulations which are
used in the Monte Carlo method. However, many optimisations of the DOProC procedures
are available and the number of computations can be effectively reduced. Anyway, if the
computation includes a finite element analysis, the required computational time can be very
long. The DOProC with most of its optimisations can be relatively easily parallelized. The
required effort to make a parallelized DOProC procedure is comparable with the effort to
make a parallel version the basic Monte Carlo procedure (without any optimisations). There
is also no need for any random number generator (neither serial nor parallel) which can make
parallelization of the DOProC method very effective. Still, there are several problems which
have to be discussed in detail. Thus the paper discusses the main principles of the DOProC
method and their parallel implementation. These approaches are discussed in detail. Results
of the parallelization are also presented.
purchase the full-text of this paper (price £22)
go to the previous paper |
|