In this paper, we introduce the design approach of integrated photonic devices by employing reinforcement learning known as attractor selection (AttSel). Here, we combined 3D FDTD with AttSel algorithm, which is based on artificial neural networks, to achieve ultra-compact and highly efficient wavelength demultiplexers with low crosstalk such as. The presented devices consist of SOI materials, which are compatible with complementary MOS technology. Consequently, the reinforcement learning is successfully applied to design smaller and superior integrated photonic devices.
Keyword: inverse design; machine learning; wavelength demultiplexing
In this manuscript, we demonstrate the design and experimental proof of an optical cloaking structure which multi-directionally conceals a perfectly electric conductor (PEC) object from an incident plane wave. The dielectric modulation around the highly reflective scattering PEC object is determined by an optimization process for multi-directional cloaking purposes.
Keyword: component; formatting; insert; style; styling
...
Our obligations and policy regarding cookies are subject to the TR Law on the Protection of Personal Data No. 6698.