Reinforcement Learning Algorithm and FDTD-based Simulation Applied to Schroeder Diffuser Design Optimization

Adam Kurowski; Bożena Kostek

doi:10.1109/access.2021.3114628

Reinforcement Learning Algorithm and FDTD-based Simulation Applied to Schroeder Diffuser Design Optimization

Abstrakt

The aim of this paper is to propose a novel approach to the algorithmic design of Schroeder acoustic diffusers employing a deep learning optimization algorithm and a fitness function based on a computer simulation of the propagation of acoustic waves. The deep learning method employed for the research is a deep policy gradient algorithm. It is used as a tool for carrying out a sequential optimization process the goal of which is to maximize the fitness function based on parameters characterizing the autocorrelation diffusion coefficient of the designed acoustic diffuser. As the autocorrelation acoustic diffusion coefficients are calculated from the polar response of a diffuser, the FDTD (finite-difference time-domain) simulation method is used to obtain a set of impulse responses necessary to calculate the polar responses of the optimized Schroeder diffusers. The results obtained from optimization based on the deep learning algorithm were compared with the outcomes of an analogous algorithm employing a genetic algorithm, and based on random selection of the Schroeder diffuser well depth pattern. We found that the best result was achieved by the deep policy gradient, as it produced outcomes, which, in terms of the provided autocorrelation diffusion coefficient, were statistically better than properties of designs provided by two other baseline approaches.

Cytowania

1

CrossRef
0

Web of Science
2

Scopus

Autorzy (2)

Adam Kurowski dr inż.
Bożena Kostek prof. dr hab. inż.

Cytuj jako

Pełna treść

pobierz publikację

pobrano 110 razy

Wersja publikacji: Accepted albo Published Version
DOI:: Cyfrowy identyfikator dokumentu elektronicznego (otwiera się w nowej karcie) 10.1109/ACCESS.2021.3114628
Licencja: otwiera się w nowej karcie

Słowa kluczowe

REINFORCEMENT LEARNING,SCHROEDER DIFFUSER OPTIMIZATION,ACOUSTIC SIMULATION,FDTD (FINITE-DIFFERENCE TIME-DOMAIN)

Informacje szczegółowe

Kategoria:: Publikacja w czasopiśmie
Typ:: artykuły w czasopismach
Opublikowano w:: IEEE Access nr 9, strony 136004 - 136017,
ISSN: 2169-3536
Język:: angielski
Rok wydania:: 2021
Opis bibliograficzny:: Kurowski A., Kostek B.: Reinforcement Learning Algorithm and FDTD-based Simulation Applied to Schroeder Diffuser Design Optimization// IEEE Access -Vol. 9, (2021), s.136004-136017
DOI:: Cyfrowy identyfikator dokumentu elektronicznego (otwiera się w nowej karcie) 10.1109/access.2021.3114628
Weryfikacja:: Politechnika Gdańska