Mitsubishi Heavy Industries, Ltd. (MHI) applies optimization technology to product design that use numerical analysis techniques such as finite element analysis. However, the optimization calculation requires a considerable number of analysis executions, so there is a problem in that obtaining an optimum solution takes time in the case of a high-cost analysis model that requires several hours for a single analysis. We applied Bayesian optimization, which is an optimization technology, in an effort to solve this problem and made it possible to calibrate the analysis model in a short time. This report presents an outline of these efforts.
