Effects of Microstrip Feed Line Width on 1 x 4 Rectangular Microstrip Antenna Array Electrical Parameters and Estimation with Artificial Neural Networks
Küçük Resim Yok
Tarih
2012
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
ELECTROMAGNETICS ACAD
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
In this study, 1 x 4 rectangular microstrip array antennas are designed at 16 GHz resonant frequency for KU Band usage on Duroid 5880 substrate that has a thickness of 0.254 mm and a dielectric constant of 2.2. Designs are simulated using HFSS v12. At these designs, by changing the feed line widths systematically, for each antenna, electrical parameters like Sit response, directivity, gain, radiation efficiency etc. are investigated for 26 array antennas in simulation media. Also, directivity and gain values are predicted with an Artificial Neural Network model. The network has four inputs as dielectric substrate thickness, resonant frequency and dielectric constant of the substrate and three outputs as directivity, gain and radiation efficiency. Multi layer perceptron structure for Artificial Neural Network model and Levenberg-Marquart learning algorithm for training the network are used. The network model is trained with 20 of 26 design data and is tested with the rest 6 ones. It is seen that the results from simulations and the neural network model are compatible with similar studies in the literature.
Açıklama
Progress in Electromagnetics Research Symposium -- MAR 27-30, 2012 -- Kuala Lumpur, MALAYSIA
Anahtar Kelimeler
Kaynak
PROCEEDINGS OF PROGRESS IN ELECTROMAGNETICS RESEARCH SYMPOSIUM (PIERS 2012)
WoS Q Değeri
N/A
