Title :
Reduced-order models in the finite element analysis
Author :
Fotyga, Grzegorz ; Nyka, Krzysztof ; Kulas, Lukasz
Author_Institution :
Dept. of Microwave & Antenna Eng., Gdansk Univ. of Technol., Gdansk, Poland
Abstract :
A novel technique of incorporating macromodels into finite element electromagnetic analysis of waveguide components is presented. Macromodels are generated by using a model order reduction algorithm (ENOR), which results in significant decrease of the number of variables, that describe the computational region. Proposed technique allows for using a few independent macromodels as well as to duplicating one macromodel in many subregions of analysed structure. Reduction of the numerical costs is obtained without sacrificing solution accuracy. Furthermore, proposed approach leads to reduction of the overall simulation time.
Keywords :
Computational modeling; Costs; Eigenvalues and eigenfunctions; Electromagnetic analysis; Electromagnetic waveguides; Equations; Finite element methods; Reduced order systems; Sparse matrices; Transmission line matrix methods; finite element method; macromodel; reduced order model;
Conference_Titel :
Microwave Radar and Wireless Communications (MIKON), 2010 18th International Conference on
Conference_Location :
Vilnius, Lithuania
Print_ISBN :
978-1-4244-5288-0