Automatic member grouping for truss optimization with multiple frequency constraints

José. P.G. Carvalho¹; Afonso C.C. Lemonge¹; Érica C.R. Carvalho¹; Patrícia H. Hallak¹; Heder S. Bernardino¹

¹ Universidade Federal de Juiz de Fora

doi:10.20906/CPS/CILAMCE2017-0162

Resumo

This paper deals with continuous sizing structural optimization problems with respect to the minimization of the masses of truss structures considering multiple natural frequencies as constraints. It can be attractive to use a reduced number of distinct cross-sectional areas minimizing costs of fabrication, transportation, storing, checking, welding, and so on. Also, it is expected a labor-saving when the structure is welded, checked and so on. In order to obtain an automatic variable linking searching for the best member grouping of the bars of the trusses analyzed in this paper, a special encoding by using cardinality constraints is considered. A CRPSO (Craziness based Particle Swarm Optimization) is the search algorithm adopted here that uses a modified velocity expression and an operator called ``craziness velocity". Some computational experiments are analyzed, presenting very interesting results providing curves of tradeoff between the optimized weights and the number of distinct cross-sectional areas used in these solutions.

Palavras-chave: Particle; Swarm; Optimization; Cardinality; Constraints; Automatic; Member; Grouping