Accurate Modeling of Antenna Structures by Means of Domain Confinement and Pyramidal Deep Neural Networks

Koziel, Slawomir; Çalık, Nurullah; Mahoutı, Peyman; Belen, Mehmet A.

doi:10.1109/tap.2021.3111299

Cited by 46 publications

(26 citation statements)

References 103 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2 One of the viable solution for this challenging problem is the utilization of data-driven surrogate models. [3][4][5][6] Datadriven surrogate models are based on approximating sampled data which allow the model to make a prediction of targeted output characteristics using the given inputs without the need of expert knowledge of the selected problem, high adaptation rate between similar problems, and high computational efficiency. 6,7 Modeling and development of data-driven models is a topic that being studied by many researchers.…”

Section: Introductionmentioning

confidence: 99%

“…[3][4][5][6] Datadriven surrogate models are based on approximating sampled data which allow the model to make a prediction of targeted output characteristics using the given inputs without the need of expert knowledge of the selected problem, high adaptation rate between similar problems, and high computational efficiency. 6,7 Modeling and development of data-driven models is a topic that being studied by many researchers. Some examples of the commonly used modeling methods for microwave applications from literature can be named as kriging, 8 neural networks, [9][10][11][12] support vector regression, [13][14][15] and deep learning.…”

Section: Introductionmentioning

confidence: 99%

“…Some examples of the commonly used modeling methods for microwave applications from literature can be named as kriging, 8 neural networks, [9][10][11][12] support vector regression, [13][14][15] and deep learning. 6,7,[16][17][18][19][20][21] Although the utilization of data-driven surrogate models is a viable technique for design and optimization of microwave components, these methods are also has challenging design problems which must be addressed with care. (I) Properly creating data set, either using three-dimensional (3D) EM simulation or experiential data, for model is one of the most crucial step.…”

Section: Introductionmentioning

confidence: 99%

“…Modeling and development of data‐driven models is a topic that being studied by many researchers. Some examples of the commonly used modeling methods for microwave applications from literature can be named as kriging, 8 neural networks, 9–12 support vector regression, 13–15 and deep learning 6,7,16–21 …”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Surrogate‐assisted optimization of 3D printed ceramic nonuniform nonplanar microstrip filter

Mahouti

Kuskonmaz

Yıldırım

2022

Micro & Optical Tech Letters

View full text Add to dashboard Cite

Herein, by using 3D printing technology and data‐driven surrogate model‐assisted optimization method, design of a ceramic material‐based nonuniform nonplanar microstrip filter is taken into the consideration in a computationally efficient and low‐cost manner. For this aim, 3D EM model of the proposed design had been used for generating training and validation data sets. Then commonly used state‐of‐the‐art regression algorithms had been used for creating accurate and fast surrogate models to create a mapping between inputs of the model and outputs of the unit element design. After that, Grey Wolf Optimization algorithm had been used for design optimization of a bandpass filter. Then, via the use of 3D printer, the optimally designed filter had been prototyped and its performance characteristics are measured. As a result, by using 3D printer technology, ceramic material, and the proposed method, design optimization of nonuniform nonplanar microstrip filter can be achieved in a computationally efficient way.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Surrogate‐assisted optimization of 3D printed ceramic nonuniform nonplanar microstrip filter

Mahouti

Kuskonmaz

Yıldırım

2022

Micro & Optical Tech Letters

View full text Add to dashboard Cite

show abstract

“…Meanwhile, a few machine learning algorithms are emerging in antenna area Burrascano et al [1999], jun Zhang et al [2003], Zhu et al [2007], Pastorino and Randazzo [2005], Xiao et al [2018], Cui et al [2020], Xiao et al [2021], Sharma et al [2020], Wu et al [2020], Nan et al [2021], Davoli et al [2021], Zhou et al [2021], Naseri and Hum [2021], Koziel et al [2021], Abdullah and Koziel [2022]. These algorithms can be summarized into two categories: 1) training surrogate models to replace the electromagnetic simulator for acceleration; 2) training inverse design models to propose new candidates.…”

Section: Introductionmentioning

confidence: 99%

A Machine Learning Generative Method for Automating Antenna Design and Optimization

Zhong¹,

Renner²,

Dou³

et al. 2022

Preprint

View full text Add to dashboard Cite

To facilitate the antenna design with the aid of computer, one of the practices in consumer electronic industry is to model and optimize antenna performances with a simplified antenna geometric scheme. Traditional antenna modeling requires profound prior knowledge of electromagnetics in order to achieve a good design which satisfies the performance specifications from both antenna and product designs. The ease of handling multi-dimensional optimization problems and the less dependence on the engineers' knowledge and experience are the key to achieve the popularity of simulation-driven antenna design and optimization for the industry. In this paper, we introduce a flexible geometric scheme with the concept of mesh network that can form any arbitrary shape by connecting different nodes. For such problems with high dimensional parameters, we propose a machine learning based generative method to assist the searching of optimal solutions. It consists of discriminators and generators. The discriminators are used to predict the performance of geometric models, and the generators to create new candidates that will pass the discriminators. Moreover, an evolutionary criterion approach is proposed for further improving the efficiency of our method. Finally, not only optimal solutions can be found, but also the well trained generators can be used to automate future antenna design and optimization. For a dual resonance antenna design with wide bandwidth, our proposed method is in par with Trust Region Framework and much better than the other mature machine learning algorithms including the widely used Genetic Algorithm and Particle Swarm Optimization. When there is no wide bandwidth requirement, it is better than Trust Region Framework.

show abstract

Advances in hybrid format‐based neuro‐transfer function techniques for parametric modeling of microwave components

Zhang

Liu

et al. 2023

Int J Numerical Modelling

View full text Add to dashboard Cite

Electromagnetic (EM) parametric modeling has become significant for EM designs of microwave devices. This paper outlines recent advances in hybrid format‐based neuro‐transfer function (TF) techniques for EM parametric modeling of microwave devices. To solve the problem of high‐sensitivity, a novel decomposition approach is discussed to develop a rational‐based neuro‐TF model of EM behavior of microwave devices. To handle the issue of non‐smoothness and discontinuity, a parametric modeling technique incorporating pole‐residue/rational and neural network hybrid transfer function (short for rational/pole‐residue hybrid neuro‐TF) of EM behavior is reviewed. This technique effectively combines rational and residue‐pole formats of the transfer functions. Compared with the residue‐pole‐based neuro‐TF modeling approach and the rational‐based neuro‐TF modeling approach, the rational/pole‐residue hybrid neuro‐TF technique allows for better accuracy in large geometrical changes and high order applications. A parametric modeling method combining neural network and polynomial‐transfer function (neuro‐PTF) is further presented as an advanced version of the rational/pole‐residue hybrid neuro‐TF method. In this approach, the pole–residue‐based transfer function and the polynomial function are used together to describe the EM responses, and can produce more accurate models, especially with large geometrical variables. Following the modeling process, trained models can be used to provide fast and accurate EM response predictions and can subsequently be used for advanced circuit and system design.

show abstract

Accurate Modeling of Antenna Structures by Means of Domain Confinement and Pyramidal Deep Neural Networks

Cited by 46 publications

References 103 publications

Surrogate‐assisted optimization of 3D printed ceramic nonuniform nonplanar microstrip filter

Surrogate‐assisted optimization of 3D printed ceramic nonuniform nonplanar microstrip filter

A Machine Learning Generative Method for Automating Antenna Design and Optimization

Advances in hybrid format‐based neuro‐transfer function techniques for parametric modeling of microwave components

Contact Info

Product

Resources

About