Shape Sensitivity Analysis and Gradient-Based Optimization of Large Structures Using MLFMA

Author

Kataja, Juhani ; Jarvenpaa, Seppo ; Toivanen, Jukka I. ; Makinen, Raino A. E. ; Yla-Oijala, Pasi

Author_Institution

Dept. of Radio Sience & Eng., Aalto Univ., Aalto, Finland

Volume

62

Issue

11

fYear

2014

fDate

Nov. 2014

Firstpage

5610

Lastpage

5618

Abstract

A fast method for computing the action of shape-differentiated electric field integral equation (EFIE) system matrix to a vector is derived exploiting the multilevel fast multipole algorithm (MLFMA). The proposed method is used in conjunction with the adjoint-variable method (AVM) to compute the shape gradient of arbitrary objective functions depending on shape of a metallic scatterer. The method is demonstrated numerically by optimizing the shape of a parabolic reflector illuminated with a half-wave dipole.

Keywords

antenna theory; electric field integral equations; gradient methods; matrix algebra; reflector antennas; sensitivity analysis; 3D antenna structure; AVM; EFIE system matrix; MLFMA; adjoint-variable method; arbitrary objective functions; gradient-based optimization; half-wave dipole; large structures; metallic scatterer; multilevel fast multipole algorithm; parabolic reflector antenna; shape gradient; shape sensitivity analysis; shape-differentiated electric field integral equation; Approximation methods; Dipole antennas; Equations; Linear programming; Optimization; Shape; Vectors; EFIE; Shape sensitivity; fast multipole method; reflector antenna; shape derivative; shape optimization;

fLanguage

English

Journal_Title

Antennas and Propagation, IEEE Transactions on

Publisher

ieee

ISSN

0018-926X

Type

jour

DOI

10.1109/TAP.2014.2347394

Filename

6877620