抄録
The goal of multi-objective optimization problems is to obtain pareto optimal solutions. However, for a designer who handles real world problems, it is important to obtain not only the optimal solutions but also the information of the design variable space. In this paper we propose an optimization method that is capable of obtaining the information of the design variable space. To obtain the information of the design variable space, an uniform global search of the design variable space is necessary. For this reason, a single objective optimization method called DIRECT has attracted much attention in this field. In this paper, we propose a multi-objective DIRECT algorithm called NSDIRECT-GA, which adapts DIRECT to multi-objective optimization and combines it with MOGA. The effectiveness of NSDIRECT-GA was examined through numerical experiments. From the conducted numerical experiments, it was found out that the accuracy and diversity of the obtained solutions are improved, and the design variable space is also searched uniformly and globaly.
