세미나

This seminar presentation aims to provide the review of existing literatures on topology optimization
using artificial neural networks (ANNs), and discuss the related future work. Topology optimization is a
promising design tool than can find the optimal structural shapes and configurations maximizing a given
design goal. Recently, huge amounts of papers have been published pursuing the application of ANNs
into structural topology optimization. In this presentation, the existing literatures are classified by the
ANN application objectives that determine the input and output of ANNs. Then, their effectives for
topology optimization is discussed in terms of practical advantages compared to the conventional
approach. Next, the future direction of topology optimization using ANNs are discussed: specifically, the
physics-informed neural networks for topology optimization will be explained.