Scholarlink

<p>This paper presents the design of a metamaterial absorber (MMA) that resonates at dual frequencies within the C-band spectrum. The dual-band absorption characteristics of the MMA are realized by employing two concentric circular rings patterned on the top surface of an FR-4 dielectric substrate, which is backed by a continuous copper ground plane. To expedite and enhance the design process, machine learning techniques are employed for optimizing the geometric parameters of the MMA. Traditionally, the design of MMAs involves extensive electromagnetic simulations and a trial-and-error approach to identify optimal dimensions, which is time-consuming and computationally expensive. In this work, the regression-based machine learning algorithms <i>k</i>-nearest neighbors (KNN) regression, random forest regression, XGBoost regression ,and CatBoost regression are implemented to model and predict the absorber performance based on its geometric parameters. Among these algorithms, CatBoost regression demonstrated the highest prediction accuracy, achieving approximately 98% accuracy in estimating the absorptivity characteristics. The optimized dimensions obtained through this model can be directly utilized to design MMAs with enhanced absorption at two targeted frequencies within the C-band. This approach not only accelerates the design process but also contributes to the development of efficient and compact electromagnetic absorbers for applications in stealth technology, sensing, and wireless communications.</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

Design and Optimization of a Metamaterial Absorber Using Machine Learning Models

  • 期刊论文

  • 发表日期：

    2025年11月7日

查看全文
  • Priyanka Garg

摘要

This paper presents the design of a metamaterial absorber (MMA) that resonates at dual frequencies within the C-band spectrum. The dual-band absorption characteristics of the MMA are realized by employing two concentric circular rings patterned on the top surface of an FR-4 dielectric substrate, which is backed by a continuous copper ground plane. To expedite and enhance the design process, machine learning techniques are employed for optimizing the geometric parameters of the MMA. Traditionally, the design of MMAs involves extensive electromagnetic simulations and a trial-and-error approach to identify optimal dimensions, which is time-consuming and computationally expensive. In this work, the regression-based machine learning algorithms k-nearest neighbors (KNN) regression, random forest regression, XGBoost regression ,and CatBoost regression are implemented to model and predict the absorber performance based on its geometric parameters. Among these algorithms, CatBoost regression demonstrated the highest prediction accuracy, achieving approximately 98% accuracy in estimating the absorptivity characteristics. The optimized dimensions obtained through this model can be directly utilized to design MMAs with enhanced absorption at two targeted frequencies within the C-band. This approach not only accelerates the design process but also contributes to the development of efficient and compact electromagnetic absorbers for applications in stealth technology, sensing, and wireless communications.