Reliable Space-Mapping Optimization Integrated With EM-Based Adjoint Sensitivities

Koziel, Slawomir; Ogurtsov, Stanislav; Bandler, J.W.; Cheng, Qingsha S.

doi:10.1109/tmtt.2013.2278148

Cited by 70 publications

(37 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For inverse problems, a physical model can be looked as an abstract system H which maps a vector of X onto a vector of Y, where X represents control variables and Y represents the needed predictions corresponding to observations [13][14][15][16][17][18] The adjoint of the PE propagation model has been widely used in atmospheric refractivity estimations [19][20][21] and ocean acoustic inversions [22][23][24], where the source information is assumed to be known and the environmental physical characteristics are estimated by matching the observed signals with predicted fields with a gradient-based minimization algorithm. For the source localization problem considered in this paper, the environmental conditions are assumed to be known.…”

Section: Adjoint Operatormentioning

confidence: 99%

Source Localization in the Duct Environment with the Adjoint of the PE Propagation Model

Zhao

2015

Atmosphere

View full text Add to dashboard Cite

Owing to the absorbing, refracting and scattering effects of the propagation medium, electromagnetic (EM) energy will degrade with the increment of propagation range, and the maximum value exists at the point of the radiating source. Employing this phenomenon, this paper introduces a novel approach to detect the location of EM transmitters in an atmospheric duct environment. Different from previous matched-field processing (MFP) methods, the proposed method determines the source location through reconstructing the forward propagation field pattern by the backward adjoint integration of the parabolic equation (PE) propagation model. With this method, the repeated computations of PE used in the MFP methods are not needed. The performance of the method is evaluated via numerical simulations, where the influences of the measurement noise and the geometry of the receiver array on the localization results are considered.

show abstract

Section: Adjoint Operatormentioning

confidence: 99%

Source Localization in the Duct Environment with the Adjoint of the PE Propagation Model

Zhao

2015

Atmosphere

View full text Add to dashboard Cite

show abstract

“…In the work of [12], [14]- [16], coarse and fine mesh EM simulations are used to enable space mapping EM optimization. Sensitivity information from EM simulations has been used to increase the effectiveness of space mapping [15], [17], [18]. The convergence speed in this case is affected by the difference between fine and coarse mesh EM simulations, and the continuity of the coarse mesh EM response w.r.t design variables.…”

mentioning

confidence: 99%

Cognition-Driven Formulation of Space Mapping for Equal-Ripple Optimization of Microwave Filters

Zhang

Feng

Gongal-Reddy

et al. 2015

IEEE Trans. Microwave Theory Techn.

Self Cite

138

View full text Add to dashboard Cite

Space mapping is a recognized method for speeding up electromagnetic (EM) optimization. Existing space-mapping approaches belong to the class of surrogate-based optimization methods. This paper proposes a cognition-driven formulation of space mapping that does not require explicit surrogates. The proposed method is applied to EM-based filter optimization. The new technique utilizes two sets of intermediate feature space parameters, including feature frequency parameters and ripple height parameters. The design variables are mapped to the feature frequency parameters, which are further mapped to the ripple height parameters. By formulating the cognition-driven optimization directly in the feature space, our method increases optimization efficiency and the ability to avoid being trapped in local minima. The technique is suitable for design of filters with equal-ripple responses. It is illustrated by two microwave filter examples.Index Terms-Cognition-driven design, computer-aided design (CAD), electromagnetic (EM) optimization, microwave filters, modeling, space mapping (SM).

show abstract

“…The second approach is classified as a geometrical approach and treats the problem implicitly by finding the center of a body approximating the feasible region, a region in the design parameter space where the design specifications are satisfied.The main obstacle in the design centering process, for both approaches, is the computational effort required for evaluating the circuit performance measures especially in the microwave circuits where many full-wave EM simulations would be required that may exhaust hours of central processing unit (CPU) time [2]. To overcome this, computationally cheap space mapping (SM) surrogate models can be employed instead of the computationally expensive high-fidelity fine model [3][4][5][6][7][8]. Several techniques for SM surrogate-based microwave circuit design centering have been proposed in the literature [2,[9][10][11][12][13].…”

mentioning

confidence: 99%

Space mapping surrogate‐based microwave circuit design centering using a new statistical technique

Hassan

Abdel-Malek

Mohamed

et al. 2015

Int J Numerical Modelling

View full text Add to dashboard Cite

Design centering is an optimal design process that seeks for the values of system parameters which maximize the probability of satisfying the design specifications (yield function). In this article, a derivative-free trust region (TR) optimization approach is combined with the generalized space mapping technique to develop a new space mapping (SM) surrogate-based statistical technique for design centering of computationally expensive microwave circuits. The TR optimization approach is well suited for expensive objective functions that have some uncertainty in their values or subject to statistical variations. The principal operation relies on building and successively updating quadratic surrogate models to be optimized instead of the objective function over TRs employing the truncated conjugate gradients. The approach constructs the initial quadratic surrogate model using fewer data points, then updates the surrogate model using a weighted least squares fitting that gives more emphasis to points close to the current center point. Integrating the TR optimization approach with the generalized space mapping technique results in a novel statistical design centering technique for the microwave circuits with a great reduction in computations and simulations needed for the yield optimization process. Design of practical microwave circuits is presented to indicate the effectiveness of the new design centering technique.

show abstract

Reliable Space-Mapping Optimization Integrated With EM-Based Adjoint Sensitivities

Cited by 70 publications

References 41 publications

Source Localization in the Duct Environment with the Adjoint of the PE Propagation Model

Source Localization in the Duct Environment with the Adjoint of the PE Propagation Model

Cognition-Driven Formulation of Space Mapping for Equal-Ripple Optimization of Microwave Filters

Space mapping surrogate‐based microwave circuit design centering using a new statistical technique

Contact Info

Product

Resources

About