構造形態の多目的最適化問題に対する優良解獲得を目指した遺伝的アルゴリズムと解空間の状況

概要

論文の詳細を見る
We proposed the genetic algorithms to obtain the decent solution with diversity of the structural morphogenesis problem. The proposed procedure is able to obtain a solution set including global optimal solution and local optimal solution with high estimation in feasible region. First of all, the mechanism of ISGA that pursues decent solution of the multiobjective optimization problem is explained. Next, the numerical experiment that applies ISGA to the multiobjective optimization problem that searches for a both high position in which two bivariate functions were used is shown. The explained mechanism is confirmed from the obtained numerical results. Based on these numerical results, ISGA is applied to the multiobjective optimization problem of the structural morphogenesis used by the past study. Finally, by using ISGA, the following items are shown from a pursuit mechanism of the decent solution with diversity and the numerical result of three analytical models. 1) The situation of the solution space can be grasped. 2) The Pareto front can be obtained with the diversity of the design variable space retained when the composition of the solution space is simple. 3) The local Pareto front can be obtained when the composition of the solution space is complex.
一般社団法人日本建築学会の論文
2008-03-25