José E. Rayas-Sánchez

Neuromodeling of microwave circuits exploiting space-mapping technology

Space mapping (SM) technology based neuromodels decrease the cost of training, improve generalization ability and reduce the complexity of the ANN topology w.r.t. classical neuromodeling. Three novel techniques are proposed to generate SM based neuromodels: space-mapped neuromodeling (SMN), frequency dependent space-mapped neuromodeling (FDSMN), and frequency-space-mapped neuromodeling (FSMN). Huber optimization is proposed to train the neuro-space-mapping (NSM). The techniques are illustrated by a microstrip right angle bend.

EM-Based Monte Carlo Analysis and Yield Prediction of Microwave Circuits Using Linear-Input Neural-Output Space Mapping

A computationally efficient method for highly accurate electromagnetics-based statistical analysis and yield estimation of RF and microwave circuits is described in this paper. The statistical analysis is realized around a space-mapped nominal solution. Our method consists of applying a constrained Broyden-based linear input space-mapping approach to design, followed by an output neural space-mapping modeling process in which not only the responses, but the design parameters and independent variable are used as inputs to the output neural network. The output neural network is trained using reduced sets of training and testing data generated around the space-mapped nominal solution. We illustrate the accuracy and efficiency of our technique through the design and statistical analysis of a classical synthetic problem and a microstrip notch filter with mitered bends

Neural space-mapping optimization for EM-based design

We propose, for the first time, neural space-mapping (NSM) optimization for electromagnetic based design. NSM optimization exploits our space-mapping (SM)-based neuromodeling techniques to efficiently approximate the mapping. A novel procedure that does not require troublesome parameter extraction to predict the next point is proposed. The initial mapping is established by performing upfront fine-model analyses at a reduced number of base points. Coarse-model sensitivities are exploited to select those base points. Huber optimization is used to train, without testing points, simple SM-based neuromodels at each NSM iteration. The technique is illustrated by a high-temperature superconducting quarter-wave parallel coupled-line microstrip filter and a bandstop microstrip filter with quarter-wave resonant open stubs.

A generalized space-mapping tableau approach to device modeling

A comprehensive framework to engineering device modeling, which we call generalized space mapping (GSM) is introduced in this paper. GSM permits many different practical implementations. As a result, the accuracy of available empirical models of microwave devices can be significantly enhanced. We present three fundamental illustrations: a basic space-mapping super model (SMSM), frequency-space-mapping super model (FSMSM) and multiple space mapping (MSM). Two variations of MSM are presented: MSM for device responses and MSM for frequency intervals. We also present novel criteria to discriminate between coarse models of the same device. The SMSM, FSMSM, and MSM concepts have been verified on several modeling problems, typically utilizing a few relevant full-wave electromagnetic simulations. This paper presents four examples: a microstrip line, a microstrip right-angle bend, a microstrip step junction, and a microstrip shaped T-junction, yielding remarkable improvement within regions of interest.

Space mapping optimization of microwave circuits exploiting surrogate models

A powerful new Aggressive Space Mapping (ASM) optimization algorithm is presented. It draws upon recent developments in both surrogate-based optimization and microwave device neuromodeling. Our surrogate formulation (new to microwave engineering) exploits, in a novel way, a linear frequency-space mapping. This is a powerful approach to severe response misalignments.

A general EM-based design procedure for single-layer substrate integrated waveguide interconnects with microstrip transitions

We propose in this work a general procedure to efficient EM-based design of single-layer SIW interconnects, including their transitions to microstrip lines. Our starting point is developed by exploiting available empirical knowledge for SIW. We propose an efficient SIW surrogate model for direct EM design optimization in two stages: first optimizing the SIW width to achieve the specified low cutoff frequency, followed by the transition optimization to reduce reflections and extend the dominant mode bandwidth. Our procedure is illustrated by designing a SIW interconnect on a standard FR4-based substrate.

A Generalized Space Mapping Tableau Approach to Device Modeling

A novel, comprehensive framework to engineering device modeling called Generalized Space Mapping (GSM) is introduced. The accuracy of available empirical models of microwave devices can be significantly enhanced by exploiting GSM. We present three fundamental illustrations: a basic Space Mapping Super Model (SMSM), a basic Frequency-Space Mapping Super Model (FSMSM) and Multiple Space Mapping (MSM). The new concept is verified on several device modeling problems, typically utilizing very few full-wave EM simulations, yielding remarkable improvement in accuracy.

Expanded space-mapping EM-based design framework exploiting preassigned parameters

We present a novel design framework for microwave circuits. We calibrate coarse models (circuit based models) to align with fine models (full-wave electromagnetic simulations) by allowing some preassigned pa- rameters (which are not used in optimization) to change in some compo- nents of the coarse model. Our expanded space-mapping design-frame- work (ESMDF) algorithm establishes a sparse mapping from optimizable to preassigned parameters. We illustrate our approach through two mi- crostrip design examples.

Space-mapping optimization of microwave circuits exploiting surrogate models

A powerful new space-mapping (SM) optimization algorithm is presented in this paper. It draws upon recent developments in both surrogate model-based optimization and modeling of microwave devices, SM optimization is formulated as a general optimization problem of a surrogate model. This model is a convex combination of a mapped coarse model and a linearized fine model. It exploits, in a novel way, a linear frequency-sensitive mapping. During the optimization iterates, the coarse and fine models are simulated at different sets of frequencies. This approach is shown to be especially powerful if a significant response shift exists. The algorithm is illustrated through the design of a capacitively loaded 10:1 impedance transformer and a double-folded stub filter. A high-temperature superconducting filter is also designed using decoupled frequency and SMs.

Neural inverse space mapping EM-optimization

For the first time, we present Neural Inverse Space Mapping (NISM) optimization for EM-based design of microwave structures. The inverse of the mapping from the fine to the coarse model parameter spaces is exploited for the first time in a Space Mapping algorithm. NISM optimization does not require: up-front EM simulations, multipoint parameter extraction or frequency mapping. The inverse of the mapping is approximated by a neural network whose generalization performance is controlled through a network growing strategy. We contrast our new algorithm with Neural Space Mapping (NSM) optimization.