Predição da intensidade do campo elétrico da onda de superfície utilizando redes neurais artificiais

AUTOR(ES)

Kátia Cristina Barbosa Loschi de Melo

DATA DE PUBLICAÇÃO

2008

RESUMO

The purpose of the current work is to create a neural network simulator capable of predicting the electric field intensity of a surface wave, from some physical characteristics of the propagation environment. For the network training, it was used data collected on the central region of Brazil, near the Brasília city, consisting of an AM radio signal transmitted at 980 kHz frequency. Initially, it is presented a brief theoretical review of the two major topics of this research: radio wave propagation and neural networks. Next, it specifies the neural network project, including details of the network architecture and description of the data treatment process used on the neural networking training and validation. Aiming for a satisfactory result, three distinct neural networks were built. In order to validate the efficiency and applicability of each designed network, its obtained results were compared with the measured electric field values for the region, and with the theoretical electric field values obtained by following the ITU-R P.368-7 prescriptions. The results obtained in this work validate the use of the designed simulator as a practical tool for new communications systems projects, including the future digital radio project.

ASSUNTO(S)

engenharia eletrica propagação de ondas de superfície redes neurais artificiais predição do campo elétrico

Documentos Relacionados

• Predição da resistencia a compressão de um cimento industrial utilizando tecnicas de redes neurais artificiais
• Estimação do volume de árvores utilizando redes neurais artificiais
• Caracterização de imagens utilizando redes neurais artificiais
• Casamento de mapas utilizando redes neurais artificiais paraconsistentes
• EMPREGABILIDADE DE REDES NEURAIS ARTIFICIAIS (RNA) NA PREDIÇÃO DA QUALIDADE DE PAINÉIS AGLOMERADOS