2021 年 20 巻 p. 114-117
This paper shows results of designing grid system using numerical simulation with genetic algorithm. Furthermore, the search processes were visualized using principal component analysis, a dimension reduction technique for deeper understanding of the search processes and evaluation of the searched area. Assuming the design of the μ 10 engine, grid parameters were optimized according to an objective function. However, the result of visualization showed that the search area was not comprehensive. In addition, verification of the population size in the genetic algorithm indicates the multimodality of the search space, which clarified the direction of future improvements.