誤差情報と深層学習を用いた新しい有限要素法

抄録

It is well known that the accuracy of solutions by the finite element method directly depends on the number of nodes used in the analysis, while employing more nodes requires much more analysis time. In this study, we propose a new finite element method that can produce accurate analysis results in a short time even with a coarse mesh. The proposed method utilizes both error information obtained by a posteriori error estimation and deep learning to improve the accuracy of analysis results. In this paper, the fundamental formulation of the proposed method is explained in detail, and its basic performance is tested through sample two-dimensional stress analyses.