Feature-Based Inverse Modeling of Nanophotonic Devices

Feature-based models can represent the mathematical link between design features and design parameters. The design features denote design performance parameters, e.g., the wavelength location and amplitude of peaks in the absorbance or transmittance response. Feature-based inverse models based on the tandem network architecture are investigated in this work. This avoids coupling a forward model with an optimizer to obtain the optimal design parameters values for a set of desired design features. The feature-based inverse models directly perform the prediction of the optimal design parameters values.