ON DESIGNING A COMPUTATIONAL EXPERIMENT SYSTEM FOR VARIOUS ENGINEERING INTERPRETATIONS OF A GLOBAL OPTIMIZATION PROBLEM

A Pevneva, G Penkov, M Bakiev
2017 unpublished
This article proposes a software system design to achieve a numerical solution of a global optimization problem. A key part of it is the classification of global search algorithms by a methodological criterion, which enables a systematic comparative analysis of specific implementations of each method within the scope of engineering interpretations of a global optimization problem. The proposed architecture accommodates modular extensions of the system, which allows designing algorithmic
more » ... tions of global search methods of various classes, taking into account the impact of parameter numerical values on computation performance. Keywords: global search algorithms, methodological classification, computational experiment, computation performance, inverse problem of mathematical modelling, comparative analysis of algorithms, expert system designing.
fatcat:gavlpidalrb7rdim6oo7mlg3om