Design of metamaterials using artificial neural networks

Freitas, Guilherme; Rego, S. L.; Vasconcelos, C. F. L.

doi:10.1109/imoc.2011.6169403

Cited by 8 publications

(4 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With time, it will be integrated into all levels of technology readiness, becoming commonplace as an assistive tool in the computational design of metamaterials. Weng, Ding, Hu, et al [266] 2020 Deep Learning Classification 10 Melo Filho, Angeli, Ophem, et al [267] 2020 ANN Inverse Design 11 Chen, Lu, Karniadakis, et al [268] 2020 Deep Learning Inverse Design 12 Kollmann, Abueidda, Koric, et al [269] 2020 Deep Learning Optimization Framework 13 Qu, Zhu, Shen, et al [270] 2020 ANN Optimization Framework 14 Lai, Amirkulova, and Gerstoft [271] 2021 CNN, GAN Inverse Design 15 Gurbuz, Kronowetter, Dietz, et al [272] 2021 GAN Inverse Design 16 Amirkulova, Tran, and Khatami [273] 2021 Deep Learning Inverse Design 17 Wu, Liu, Jahanshahi, et al [274] 2021 ANN Inverse Design 18 Shah, Zhuo, Lai, et al [275] 2021 RL Optimization Framework 19 Tran, Amirkulova, and Khatami [276] 2022 ANN Inverse Design 20 Wiest, Seepersad, and Haberman [277] 2022 GNN Inverse Design 21 Amirkulova, Zhou, Abbas, et al [278] 2022 Deep Learning Inverse Design 22 Tran, Khatami, and Amirkulova [279] 2022 CNN Inverse Design 23 Li, Chen, Li, et al [280] 2023 CNN Inverse Design 24 Li, Chen, Li, et al [281] 2023 Deep Learning Inverse Design 25 Wang, Chen, Xu, et al [282] 2023 ANN Inverse Design Application field: Electromagnetics 26 Jiang, Xiao, Liu, et al [283] 2010 ANN and scaled conjugate Surrogate model gradient 27 Freitas, Rêgo, and Vasconcelos [140] 2011 ANN Surrogate model 28 Vasconcelos, Rêgo, and Cruz [139] 2012 ANN Surrogate model 29 Sarmah, Sarma, and Baruah [172] 2015 ANN optimization framework 30 Saha and Maity [138] 2016 ANN Surrogate model 31 Nanda, Sahu, and Mishra [108] 2019 ANN Inverse design 32 An, Fowler, Shalaginov, et al [144] 2019 ANN Surrogate model 33 Yuze, Hai, and Qinglin [157] 2019 CNN Classification and clustering 34 Liu, Zhang, and Cui [156] 2019 CNN optimization framework 35 Hodge, Mishra, and Zaghloul [284] 2019 DC-GAN Inverse design 36 Hodge, Mishra, and Zaghloul…”

Section: The Causal Relationship Problemmentioning

confidence: 99%

Machine Intelligence in Metamaterials Design

Cerniauskas,

Sadia,

Alam

2023

Preprint

View full text Add to dashboard Cite

Machine intelligence continues to rise in popularity as an aid to the design and discovery of novel metamaterials. The properties of metamaterials are essentially controllable via their architectures and until recently, the design process has relied on a combination of trial-and-error and physics-based methods for optimization. These processes can be time-consuming and challenging, especially if the design space for metamaterial optimization is explored thoroughly. Artificial intelligence (AI) and machine learning (ML) can be used to overcome challenges like these as pre-processed massive metamaterial datasets can be used to very accurately train appropriate models. The models can be broad, describing properties, structure, and function at numerous levels of hierarchy, using relevant inputted knowledge. Here, we present a comprehensive review of the literature where state-of-the-art machine intelligence is used for the design, discovery and development of metamaterials. In this review, individual approaches are categorized based on methodology and application. We further present machine intelligence trends over a wide range of metamaterial design problems including: acoustics, photonics, plasmonics, mechanics, and more. Finally, we identify and discuss recent research directions and highlight current gaps in knowledge. KeywordsMetamaterials • Artificial intelligence • Machine learning • Neural Networks • Evolutionary algorithms • Surrogate modelling • Optimization 42 solely related to the properties of material constituents. 43 Properties can therefore be controlled and manipulated by 44 changing metamaterial 'unit cell' topology and while the 45 bulk properties of the constituent materials have influence, 46 these properties are often not the sole dominating variable 47 used in designing metamaterials. When designing with 48 respect to topology, the properties of metamaterials can 49 be customized within a very large design space, and this 50 is one of the reasons for the exponential growth in meta-51 materials R&D across the breadth of modern engineering.

show abstract

Section: The Causal Relationship Problemmentioning

confidence: 99%

Machine Intelligence in Metamaterials Design

Cerniauskas,

Sadia,

Alam

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Neural networks have been used to solve the design, optimization and prediction problems of electromagnetism in some early works [45][46][47], but the model capability and performance were limited, largely due to the simple model structure and the lack of data. More recent works sought to deal with the inverse design by DL under various scenarios, like plasmonic waveguide [48], optical power splitter [49], plasmonic metamaterials [50], chiral metamaterials [51] and nanophotonic particles [52].…”

Section: Introductionmentioning

confidence: 99%

Probabilistic Representation and Inverse Design of Metamaterials Based on a Deep Generative Model with Semi‐Supervised Learning Strategy

et al. 2019

View full text Add to dashboard Cite

mostly relies on physics-inspired methods, resorting to human knowledge such as physical insights revealed by simplified analytical modeling, similar experience transferred from previous practice, and intuition obtained by scientific reasoning. For example, many meta-atoms inherited traditional antenna designs with geometries like rectangle, [4] cross, [5] bowtie, [6] V-shape, [7] H-shape, [8] and so on, whose first-order response is approximated by electrical dipole resonance with relevant scaling effect. [9] Some other designs guided by physical intuition include ring-like structures that exhibit strong magnetic resonances induced by the incident magnetic field, [10][11][12] dielectric building blocks that can induce both electric and magnetic resonances leading to better control of the phase of the scattered light, [13][14][15] or the spectra line-shape tailoring by introducing coupling among different resonant modes. [16][17][18] Despite the exciting results obtained by these physics-inspired designs, this methodology basically relies on a trial-and-error process, usually involving numerical methods like finite-difference-time-domain (FDTD) or finite element method (FEM) to iteratively solve Maxwell's equations. The low efficiency and thus limited exploration of the design varieties tend to easily omit the optimal solution. The inverse design approaches start from the opposite end, and try to optimize certain objective functions describing the desired performance. [19,20] Common approaches for inverse problems include genetic algorithm, [21] level set methods, [22] and topology optimization, [23] which, however, are still stochastic searching algorithms that are time-consuming and deteriorate rapidly as the design space grows. Different from numerical calculations, data-driven methods based on machine learning (ML) solve the optimization problem from statistical perspectives, so that the solution to optimize a target can be approximately generalized from numerous design examples. With the rapidly accumulated data and thus booming of deep learning (DL), the state-of-theart in many research domains, such as speech recognition, [24] computer vision, [25,26] natural language processing, [27] and decision making, [28] has been pushed far beyond conventional methods. Deep neural networks simulate biological signal processing that allow computational models to learn multiple The research of metamaterials has achieved enormous success in the manipulation of light in a prescribed manner using delicately designed subwavelength structures, so-called meta-atoms. Even though modern numerical methods allow for the accurate calculation of the optical response of complex structures, the inverse design of metamaterials, which aims to retrieve the optimal structure according to given requirements, is still a challenging task owing to the nonintuitive and nonunique relationship between physical structures and optical responses. To better unveil this implicit relationship and thus facilitate metamaterial designs, it is proposed ...

show abstract

“…Recently, propelled by its success in computer vision and natural language processing, DL has emerged as a revolutionary and powerful methodology in many other research fields such as materials science, chemistry, particle physics, quantum mechanics, , and microscopy . As the most widely used component in a DL architecture, neural networks have been applied to solve some design and prediction problems of electromagnetism − but with limited success largely because of the shallow structure and thus poor representation capability. Some very recent works have proposed deep neural networks to model nanophotonic structures, − which, however, are mainly constructed by stacking several fully connected layers and, therefore, can only deal with simple structure designs with limited optical responses.…”

mentioning

confidence: 99%

Deep-Learning-Enabled On-Demand Design of Chiral Metamaterials

2018

View full text Add to dashboard Cite

Deep-learning framework has significantly impelled the development of modern machine learning technology by continuously pushing the limit of traditional recognition and processing of images, speech, and videos. In the meantime, it starts to penetrate other disciplines, such as biology, genetics, materials science, and physics. Here, we report a deep-learning-based model, comprising two bidirectional neural networks assembled by a partial stacking strategy, to automatically design and optimize three-dimensional chiral metamaterials with strong chiroptical responses at predesignated wavelengths. The model can help to discover the intricate, nonintuitive relationship between a metamaterial structure and its optical responses from a number of training examples, which circumvents the time-consuming, case-by-case numerical simulations in conventional metamaterial designs. This approach not only realizes the forward prediction of optical performance much more accurately and efficiently but also enables one to inversely retrieve designs from given requirements. Our results demonstrate that such a data-driven model can be applied as a very powerful tool in studying complicated light-matter interactions and accelerating the on-demand design of nanophotonic devices, systems, and architectures for real world applications.

show abstract

Design of metamaterials using artificial neural networks

Cited by 8 publications

References 9 publications

Machine Intelligence in Metamaterials Design

Machine Intelligence in Metamaterials Design

Probabilistic Representation and Inverse Design of Metamaterials Based on a Deep Generative Model with Semi‐Supervised Learning Strategy

Deep-Learning-Enabled On-Demand Design of Chiral Metamaterials

Contact Info

Product

Resources

About