2010
DOI: 10.1002/jnm.743
|View full text |Cite
|
Sign up to set email alerts
|

Recent advances in space‐mapping‐based modeling of microwave devices

Abstract: SUMMARYWe review the latest developments in space-mapping-based modeling techniques with applications in microwave engineering. We discuss the two techniques that utilize a combination of standard space mapping and function approximation methodologies, in particular fuzzy systems and support vector regression (SVR). In both cases, the initial space-mapping model is enhanced by an additional term that approximates the differences between the fine model and the initial space-mapping surrogate. We compare the sta… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
5

Citation Types

0
19
0

Year Published

2011
2011
2016
2016

Publication Types

Select...
7
1

Relationship

7
1

Authors

Journals

citations
Cited by 25 publications
(19 citation statements)
references
References 61 publications
0
19
0
Order By: Relevance
“…Perhaps the most popular type of technique of this kind is space mapping (SM) [22][23][24][25], where the surrogate is constructed by means of suitable correction of an underlying lowfidelity (or so-called coarse) model. The bottleneck of space mapping in terms of antenna modeling is the lack of fast coarse models, because low-fidelity antenna representations are normally obtained through coarse-discretization EM simulations, the cost of which cannot be neglected.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…Perhaps the most popular type of technique of this kind is space mapping (SM) [22][23][24][25], where the surrogate is constructed by means of suitable correction of an underlying lowfidelity (or so-called coarse) model. The bottleneck of space mapping in terms of antenna modeling is the lack of fast coarse models, because low-fidelity antenna representations are normally obtained through coarse-discretization EM simulations, the cost of which cannot be neglected.…”
Section: Introductionmentioning
confidence: 99%
“…Another issue with SM is fixed number of extractable parameters which limits the model flexibility. This particular difficulty can be alleviated, to some extent by SM enhancement through fuzzy systems [26], radial-basis functions [23], or Kriging [27]. The problem of excessive number of training samples necessary to establish a reliable surrogate can be partially addressed by modeling methods that rely on appropriately extracted response features (e.g., shape-preserving response prediction [28], or feature-based modeling [29], however, these methods impose relatively strong assumptions on the response shapes of the structures under consideration so their applicability is limited to certain types of devices [29].…”
Section: Introductionmentioning
confidence: 99%
“…Another issue with SM is fixed number of extractable parameters which limits the model flexibility. This particular difficulty can be alleviated, to some extent by SM enhancement through fuzzy systems [21], radial-basis functions [18], or Kriging [22].…”
Section: Introductionmentioning
confidence: 99%
“…The most popular technique of this kind in microwave engineering is space mapping (SM) [20]- [24]. In SM, the surrogate is constructed by means of a suitable correction of a low-fidelity (coarse) model of the microwave structure in question (high-fidelity or fine model), e.g., some auxiliary mappings applied to a circuit equivalent "reshape" the parameter space and/or response of the circuit [22]. The enhancement of the low-fidelity model is typically realized through suitable analytical formulas, which allows the surrogate model to be almost as computationally cheap as the coarse model.…”
Section: Introductionmentioning
confidence: 99%
“…The enhancement of the low-fidelity model is typically realized through suitable analytical formulas, which allows the surrogate model to be almost as computationally cheap as the coarse model. SM surrogate model identification is normally realized using a nonlinear parameter extraction process [22]. Due to the fact that the underlying coarse model embeds some knowledge about the structure under consideration, the accuracy of the SM surrogate is considerably better than the accuracy of possible function approximation models using a comparable amount of fine model data [22].…”
Section: Introductionmentioning
confidence: 99%