Abstract
AbstractMachine learning offers the potential to revolutionize the inverse design of complex nanophotonic components. Here, we propose a novel variant of this formalism specifically suited for the design of resonant nanophotonic components. Typically, the first step of an inverse design process based on machine learning is training a neural network to approximate the non-linear mapping from a set of input parameters to a given optical system’s features. The second step starts from the desired features, e.g. a transmission spectrum, and propagates back through the trained network to find the optimal input parameters. For resonant systems, this second step corresponds to a gradient descent in a highly oscillatory loss landscape. As a result, the algorithm often converges into a local minimum. We significantly improve this method’s efficiency by adding the Fourier transform of the desired spectrum to the optimization procedure. We demonstrate our method by retrieving the optimal design parameters for desired transmission and reflection spectra of Fabry–Pérot resonators and Bragg reflectors, two canonical optical components whose functionality is based on wave interference. Our results can be extended to the optimization of more complex nanophotonic components interacting with structured incident fields.
Highlights
Inverse design of optical components is the process of calculating the material properties that yield a specific optical response
We demonstrate our method by retrieving the optimal design parameters for desired transmission and reflection spectra of Fabry–Pérot resonators and Bragg reflectors, two canonical optical components whose functionality is based on wave interference
Finding the global minimum in a parameter space where there are many local minima is a significant problem in inverse design
Summary
Inverse design of optical components is the process of calculating the material properties that yield a specific optical response. A prevalent method in density topology optimization is the adjoint method This technique provides a way to efficiently compute gradients of a loss function with respect to design parameters, as was demonstrated in the design of demultiplexers that separate light of different wavelengths [5, 6]. A neural network is trained to predict the particles’ scattering cross-section as a function of each shell’s thicknesses This network is trained using a supervised learning algorithm. Et al improved this technique of inverse design of nanoparticles in 2019 [11] by adapting the loss function of the neural network to learn the thickness as well as the material of each layer. A problem arises when applying these inverse design methods to optical systems with strong resonances These resonances are often based on the interference of many partial waves. We propose a solution to this problem, based on Fourier analysis
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
