Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : October 18, 2024 - October 20, 2024
In the design of phononic crystals (PnCs), many studies have focused on maximizing the band gap (BG) size through various optimization methods such as topology optimization, Monte Carlo simulations, and other optimization algorithms. On the other hand, it is difficult to find materials properties and structures of PnCs with desired BG frequency and size as an "inverse problem approach". Recently, new approaches to the optimization problems based on the development of artificial intelligence technology have been attracting much attention. In this study, we utilize a deep learning model in combination with a genetic algorithm (NSGA-II) to solve the inverse problem, developing a methodology to identify the optimal material properties and structural parameters of PnCs that achieve specific BG frequency and size requirements.
