Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2015-09-28T20:12:48Z No. of bitstreams: 1 RafaelNunesDeAlmeidaPrado_TESE.pdf: 3641446 bytes, checksum: 85d829efbefde8d41eb4b62067431aa5 (MD5) / Approved for entry into archive by Elisangela Moura (lilaalves@gmail.com) on 2015-10-01T13:47:06Z (GMT) No. of bitstreams: 1 RafaelNunesDeAlmeidaPrado_TESE.pdf: 3641446 bytes, checksum: 85d829efbefde8d41eb4b62067431aa5 (MD5) / Made available in DSpace on 2015-10-01T13:47:06Z (GMT). No. of bitstreams: 1 RafaelNunesDeAlmeidaPrado_TESE.pdf: 3641446 bytes, checksum: 85d829efbefde8d41eb4b62067431aa5 (MD5) Previous issue date: 2014-06-06 / Os m?todos de Intelig?ncia Computacional v?m adquirindo espa?o nas aplica??es industriais devido a sua capacidade de solu??o de problemas na engenharia, conseq?entemente, os sistemas embarcados acompanham a tend?ncia do uso das ferramentas computacionais inteligentes de forma embarcada em m?quinas. Existem diversos trabalhos na ?rea de sistemas embarcados e sistemas inteligentes puros ou h?bridos, por?m, s?o poucos os que uniram ambas as ?reas em um s? projeto. O objetivo deste trabalho foi implementar um sistema fuzzy neural adaptativo em hardware com treinamento online para embarque em Field Programable Gate Array - FPGA. A adapta??o do sistema pode ocorrer durante a execu??o de uma determinada aplica??o, visando melhora do desempenho de forma online. A arquitetura do sistema ? modular, possibilitando a configura??o de v?rias topologias de redes fuzzy neurais com treinamento online. Verificou-se que o sistema proposto obteve desempenho satisfat?rio quando aplicado a problemas de interpola??o, classifica??o de padr?es e a problemas industriais. Diante dos resultados dos experimentos foram discutidas as vantagens e desvantagens do treinamento online em hardware ser realizado de forma paralela e serializada, esta ?ltima forma proporcionou economia na ?rea utilizada de FPGA, j? a forma de treinamento paralelo demonstrou alto desempenho e reduzido tempo de processamento. O trabalho utilizou ferramentas de desenvolvimento dispon?veis para circuitos FPGA. / Computational Intelligence Methods have been expanding to industrial applications motivated by their ability to solve problems in engineering. Therefore, the embedded systems follow the same idea of using computational intelligence tools embedded on machines. There are several works in the area of embedded systems and intelligent systems. However, there are a few papers that have joined both areas. The aim of this study was to implement an adaptive fuzzy neural hardware with online training embedded on Field Programmable Gate Array ? FPGA. The system adaptation can occur during the execution of a given application, aiming online performance improvement. The proposed system architecture is modular, allowing different configurations of fuzzy neural network topologies with online training. The proposed system was applied to: mathematical function interpolation, pattern classification and selfcompensation of industrial sensors. The proposed system achieves satisfactory performance in both tasks. The experiments results shows the advantages and disadvantages of online training in hardware when performed in parallel and sequentially ways. The sequentially training method provides economy in FPGA area, however, increases the complexity of architecture actions. The parallel training method achieves high performance and reduced processing time, the pipeline technique is used to increase the proposed architecture performance. The study development was based on available tools for FPGA circuits.