分枝拡散型PSOの開発

抄録

There are more and more issues, requirements and complexities that engineering design problems must take into account in industrial applications. Therefore, the number of behavior and design variables tends to increase, making optimization problem harder. PSO was proposed and has been widely used for its simplicity and capability to solve problems, even though their convexity is not guaranteed. However, its effectiveness is limited to a relatively small number of design variables. In this study, we propose the following four ideas to extend its efficacy for the number of design variables - adjustments for PSO parameters, a large explosion, a small explosion, and a genetic estimation technique. The explosion is kind of mutation in GA. The genetic estimation uses a chromosome called Hoxgene to identify variable dependencies, enabling PSO to solve the problem with a smaller number of variables. The results of experiments on benchmark problems showed the highly accurate solutions were stably obtained even for problems with a large number of variables using the proposed method.