Estudio de optimización aplicando algoritmos genéticos en estructuras paramétricas tipo celosia de acero.

Abstract

In the present work, a matrix analysis was made in two-dimensional lattice-type steel structures with programs in the M language on the MATLAB platform, then an automation of the algorithm was made to avoid manual data entry processes in order to achieve, in the future, a topological and shape optimization, these algorithms were validated with exercises in the SAP2000 program. Afterwards, a sections optimization was made, creating algorithms for this procedure, genetic algorithms were used for the optimization in which two selection method were used, by tournament and roulette, in order to determine the best methods; A Pratt type and a circular structure parameterization was also made, this way manual data entry of all the geometry was avoided, only entering the analyze structure parameters, this is very important for the shape and topology optimization. Finally, the shape, topology and sections optimization was made using genetic algorithms with the tournament method, this optimization was separated into two parts, one of shape and topology and another of sections in order to reduce computational time. All these optimized algorithms that were created were validated with previous studies to determine the effectiveness of the created algorithm.