| Summary: | Made available in DSpace on 2014-12-17T14:55:56Z (GMT). No. of bitstreams: 1
MarcioGP.pdf: 1534925 bytes, checksum: d1c777b1e76b23d509caeb3258a0aa97 (MD5)
Previous issue date: 2006-06-19 === This dissertation contributes for the development of methodologies through feed forward artificial neural networks for microwave and optical devices modeling. A bibliographical revision on the applications of neuro-computational techniques in the areas of microwave/optical engineering was carried through. Characteristics of networks MLP, RBF and SFNN, as well as the strategies of supervised learning had been presented. Adjustment expressions of the networks free parameters above cited had been deduced from the gradient method. Conventional method EM-ANN was applied in the modeling of microwave passive devices and optical amplifiers. For this, they had been proposals modular configurations based in networks SFNN and RBF/MLP objectifying a bigger capacity of models generalization. As for the training of the used networks, the Rprop algorithm was applied. All the algorithms used in the attainment of the models of this dissertation had been implemented in Matlab === Esta disserta??o contribui para o desenvolvimento de metodologias atrav?s de redes neurais artificiais de alimenta??o direta para a modelagem de dispositivos de microondas e ?pticos. Uma revis?o bibliogr?fica sobre as aplica??es de t?cnicas neuro-computacionais na ?reas de engenharia de microondas e ?ptica foi realizada. As caracter?sticas das redes MLP, RBF e SFNN, bem como as estrat?gias de aprendizado supervisionado foram apresentadas. As express?es de ajuste dos par?metros livres das redes acima citadas foram deduzidas a partir do m?todo do gradiente. O m?todo convencional EM-ANN foi aplicado na modelagem de dispositivos passivos de microondas e amplificadores ?pticos. Para isto, foram propostas configura??es modulares baseadas em redes SFNN e RBF/MLP objetivando uma maior capacidade de generaliza??o dos modelos. No que se refere ao treinamento das redes utilizadas, o algoritmo Rprop foi aplicado. Todos os algoritmos utilizados na obten??o dos modelos desta disserta??o foram implementados em Matlab
|
|---|
