Inverse modeling of pseudo-interdigital bandpass filters using artificial neural networks
Küçük Resim Yok
Tarih
2013
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
A neural network trained to model original EM problems can be called as the forward model where the model inputs are physical or geometrical parameters and outputs are electrical parameters. Conversely neural network techniques are applicable to inverse modeling of microwave circuit design. In opposition to conventional statistical electromagnetic signal processing applications, inverse modeling techniques acquire electrical parameters as model input and geometrical properties as the output. Pseudo-interdigital (PID) bandpass microstrip filters offer compact and planar solutions to wide bandwidth filtering applications. They avoid the through vias required for short circuiting in conventional interdigital filters. Miniaturized microstrip bandpass filters are in demand for systems requiring small size and light weight. The coupling of the resonators in filter design must be adjusted using EM simulators. There are no analytical or numerical methods proposed for accurate determination of resonator spacing. In this study, the inverse modeling is applied to accurately determine the resonators' locations consistent with desired filter specifications.
Açıklama
Progress in Electromagnetics Research Symposium, PIERS 2013 Stockholm -- 12 August 2013 through 15 August 2013 -- Stockholm -- 99730
Anahtar Kelimeler
Kaynak
Progress in Electromagnetics Research Symposium
WoS Q Değeri
N/A
Scopus Q Değeri
N/A
