Framework for Expediting Discovery of Optimal Solutions with Blackbox Algorithms in Non-Topology Photonic Inverse Design

Abstract

Photonic inverse design has emerged as an indispensable engineering tool for complex optical systems. In many instances it is important to optimize for both material and geometry configurations, which results in complex nonsmooth search spaces with multiple local minima. Finding solutions approaching the global optimum may present a computationally intractable task. Here, we develop a framework that allows expediting the search of solutions close to a global optimum on complex optimization spaces with blackbox algorithms for nontopology inverse design problems in photonics. We study the way representative blackbox optimization algorithms work, including a genetic algorithm (GA), particle swarm optimization (PSO), simulated annealing (SA), and a mesh adaptive direct search (NOMAD). We then propose and utilize a two-step approach that identifies best performance algorithms on highly irregular and nonsmooth search spaces with 60+ optimization parameters. We reveal a connection between the search space complexity and algorithm performance and find that PSO and NOMAD consistently deliver better performance for mixed integer problems encountered in photonic inverse design, particularly with the account of material combinations. Our results differ from a commonly anticipated advantage of GA. Our findings will foster a more efficient design of photonic systems with optimal performance.

Deep learning aids simultaneous structure–material design discovery: a case study on designing phase change material metasurfaces

“…The use of metaheuristics (genetic algorithm, differential evolution, particle swarm optimization, etc.) in a material-geometry landscape, as reported in the work by Digani et al 34 in the context of stacked materials, has been proven to be computationally burdensome. In these population-based optimizers, as a rule of thumb, the population size is increased in proportion to the dimensionality of the solution space.…”

“…Using the electromagnetic (EM) solver to analyze the fitness of a large population makes the inverse design computationally inefficient. 34 Data-driven approaches that involve "learning" the empirical relation between the structurematerial and its optical response are being extensively studied by the nanophotonics community. Machine learning, specifically deep learning (DL), 35,36 is being increasingly investigated [37][38][39][40][41][42] for faster and computationally efficient inverse design problems; these includee nanoresonators, 43,44 plasmonics, [45][46][47] metasurfaces/metamaterials, 47,48 topological photonics, 45,[48][49][50] and integrated photonics.…”

“…The use of metaheuristics (genetic algorithm, differential evolution, particle swarm optimization, etc.) in a material-geometry landscape, as reported in the work by Digani et al 34 . in the context of stacked materials, has been proven to be computationally burdensome.…”

An Integrated Optical Circuit Architecture for Inverse-Designed Silicon Photonic Components

“…We can identify specific developments that are relevant to the work reported in this paper. These include the computational techniques of inverse design that are assisted by deep neural networks [ 11 ], deep learning [ 12 , 13 ], physics-informed machine learning [ 14 ], black box algorithms [ 15 ], integral equation methods [ 16 ], and linkage tree genetic algorithms [ 17 ]. Additional computational methods include photonic emulation [ 18 ], the deep adjoint approach [ 19 ], phase-injected topology optimization [ 20 ], and boundary integral methods [ 21 ].…”

Increasing the stability margins using multi-pattern metasails and multi-modal laser beams

“…Moreover, one of the crucial requirements for successful photonic propulsion is an ultimate design of lightsail to accomplish a stable beam-riding while minimizing acceleration time along with proper thermal management via radiative cooling 3 , 6 , 20 , 21 . Successful beam-riding requires the generation of sufficient restoring force and torques against the possible displacement and tilt of the beam with respect to the center of the sail 1 , 2 , 22 – 28 . Previously, it has been well understood that flat macroscopic structures are unequivocally unstable and any slight misalignment or displacement will cause the sail to deviate away from the center of the beam immediately.…”