サロゲートモデルによるシミュレーションの設計初期段階への適用検討

抄録

With the recent development of data science, by utilizing surrogate models using simulation results as learning data, simulation results can be reproduced with certain accuracy in a very short time. By using this feature, it will be possible to use surrogate models in the initial stage of design and/or examination of manufacturing conditions. both of which have been difficult to apply conventional simulations due to the time required for calculation. This paper investigates the potential of surrogate model in application to these kind of problems by a case study of electroplating processes..