Scholarlink

<p>The inverse design of metasurfaces is challenging due to complex interdependencies between structural parameters and electromagnetic responses, making traditional optimization methods computationally expensive and often suboptimal. This study employs a hybrid quantum-classical machine learning approach, known as the Latent Style-based Quantum GAN (LaSt-QGAN), to perform the inverse design of metasurfaces. This method integrates a Variational Autoencoder (VAE) with a Quantum Generative Adversarial Network (QGAN) to enhance the optimization of metasurface designs aimed at achieving narrow-band absorption. The proposed method results in a 6-fold reduction in training time and a 40-fold decrease in data requirements compared to traditional GAN-based approaches in both simulator and quantum hardware. The produced metasurface designs demonstrated a higher fidelity in relation to the target absorption spectra compared to the classical GAN method considered. Importantly, the proposed model generates metasurface designs with higher Q-factor (Q=<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(10^4\)</EquationSource> </InlineEquation>) than those trained on (Q=<InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(10^3\)</EquationSource> </InlineEquation>). Additionally, the integration of a material lookup table facilitates manufacturability by allowing the substitution of predicted material properties with viable alternatives, while preserving spectral response.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Hybrid Quantum-Classical Inverse Design of Metasurfaces for Tailored High Q-factor Response

  • 期刊论文

  • 发表日期：

    2026年3月26日

查看全文
  • Sreeraj Rajan Warrier,
  • Jayasri Dontabhaktuni

摘要

The inverse design of metasurfaces is challenging due to complex interdependencies between structural parameters and electromagnetic responses, making traditional optimization methods computationally expensive and often suboptimal. This study employs a hybrid quantum-classical machine learning approach, known as the Latent Style-based Quantum GAN (LaSt-QGAN), to perform the inverse design of metasurfaces. This method integrates a Variational Autoencoder (VAE) with a Quantum Generative Adversarial Network (QGAN) to enhance the optimization of metasurface designs aimed at achieving narrow-band absorption. The proposed method results in a 6-fold reduction in training time and a 40-fold decrease in data requirements compared to traditional GAN-based approaches in both simulator and quantum hardware. The produced metasurface designs demonstrated a higher fidelity in relation to the target absorption spectra compared to the classical GAN method considered. Importantly, the proposed model generates metasurface designs with higher Q-factor (Q= \(10^4\) ) than those trained on (Q= \(10^3\) ). Additionally, the integration of a material lookup table facilitates manufacturability by allowing the substitution of predicted material properties with viable alternatives, while preserving spectral response.