Abstract
The authors previously developed the SA/GA hybrid optimization method using a procedure of mesh size change, which is very effective method for the simultaneous optimization of the topology and geometry of a two-dimensional framed structure. However, the region searching the optimum solutions can vary depending on the types of the search individuals generated stochastically in the early stages of the SA and GA operations. Consequently, it is necessary to conduct multiple trial calculations in order to obtain the Pareto-optimal solutions globally and accurately. In the present study, therefore, the two calculation techniques are newly introduced into the previous method in an attempt to improve the computation efficiency. One is aimed at generating search individuals with a great diversity of topologies during the initial stage of the iterative calculation, and the other is aimed at preventing the search individuals with the potentiality of generating the non-dominate solutions from disappearing by competition between search individuals. The validity of the present algorithm is confirmed by the results of numerical computations.
