[10a-W321-1~10] 23.1 Joint Session N "Informatics"

△ [10a-W321-7] Constraint and batch Bayesian optimization of grinding condition of SiC wafer

〇(M1C)Keichi Osada^1,2, Yosuke Tsunooka^1,2,3, Kiyoshi Narita⁴, Haruhiko Koizumi², Kentaro Kutsukake⁵, Shunta Harada^1,2, Miho Tagawa^1,2, Toru Ujihara^1,2,3 (1.Grad. School Eng., Nagoya Univ., 2.IMaSS, Nagoya Univ., 3.GaN-OIL, AIST, 4.Nitolex, 5.AIP, RIKEN)

Keywords:Bayesian optimization, SiC, optimization of process

Bayesian optimization is suitable for the optimization of material process. In the optimization of some material process, the most efficient number of the data is multiple, or the efficiency should be improved under the condition which demands some quality standard. By applying constraint and batch Bayesian optimization to grinding process of SiC wafer, this study results in the condition which decreased the damage to near the surface of SiC wafer after grinding. After now on, it’s expected that we get the condition whose grinding speed is faster than previous one, under the condition which has the damage is less than the previous one has.