Presentation Information

9:40 AM - 10:00 AM JST(12:40 AM - 1:00 AM UTC)

[1-84]Shape optimization of flow around a two-dimensional body using a PINN with boundary conditions imposed as hard constraints

*Ayato Hirayama¹, Sho Watanabe², Yusuke Yugeta¹, Takumi Endo¹, Yosuke Hasegawa³ (1. Department of Mechanical Engineering, Graduate School of Engineering, The University of Tokyo, 2. Development Division of Komatsu Ltd, 3. Institute of Industrial Science, the University of Tokyo)

Keywords:

Physics-Informed Neural Network (PINN),Machine Learning,Shape optimization,Flow estimation,Hard constraints

Back to Session information