Metasurface parameter optimization of Fano resonance based on a BP-PSO algorithm.

Area of Science:

• Photonics and Nanotechnology
• Metamaterials and Plasmonics
• Optical Engineering

Background:

• All-dielectric metasurfaces offer unique optical properties.
• Fano resonance in metasurfaces enables sharp spectral features.
• Optimizing structural parameters is crucial for enhancing metasurface performance.

Purpose of the Study:

• To propose an all-dielectric metasurface design.
• To analyze Fano resonance and its underlying mechanisms.
• To optimize metasurface parameters for superior spectral performance.

Main Methods:

• Numerical simulation of transmission spectrum.
• Analysis of Fano resonance using multipole coupling theory.
• Development of a mathematical model with back propagation (BP) neural networks.
• Optimization using genetic algorithm, sparrow search algorithm, and particle swarm optimization (PSO).

Main Results:

• Fano resonance was observed in the transmission spectrum.
• A BP neural network model accurately correlated structural parameters with spectral performance.
• PSO optimization resulted in a threefold increase in quality factor (3805) and near 100% modulation depth.

Conclusions:

• The proposed metasurface design exhibits enhanced optical properties.
• Advanced optimization algorithms, particularly PSO, are effective for tuning metasurface performance.
• This work presents a novel approach for designing and optimizing optical micro-nano structures.