• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Master's Dissertation
DOI
https://doi.org/10.11606/D.55.2018.tde-03062015-142611
Document
Author
Full name
Richard Aderbal Gonçalves
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2002
Supervisor
Committee
Lins, Rafael Dueire
Ordoñez, Edward David Moreno
Title in Portuguese
Architect-R: uma ferramenta para o desenvolvimento de robôs móveis configuráveis
Keywords in Portuguese
Não disponível
Abstract in Portuguese
Nos últimos anos o interesse por robôs, em especial por robôs móveis, tem crescido muito. Grandes avanços foram obtidos nessa área. Contudo a tarefa de programar um robô móvel ainda se mostra uma tarefa extrema,mente árdua tanto em termos da complexidade quanto do tempo desprendido. Algumas novas linguagens próprias para a programação de robôs têm sido propostas pelo meio académico, tais como a CES desenvolvida pelo Institute of Robotics da CMU-USA. Outra tarefa essencial para que os robôs móveis tenham um futuro ainda mais promissor é a criação de hardware dedicado para acelerar a execução dos algoritmos que controlam o robô de forma que ele possa tomar decisões mais complexas de maneira mais rápida. Essa tarefa envolve o conhecimento profundo dos algoritmos utilizados pelos mesmos além de conhecimentos sólidos no projeto e desenvolvimento de hardware especializado. Na tentativa de amenizar essas dificuldades, será proposta neste trabalho uma ferramenta para o desenvolvimento de robôs móveis reconfiguráveis, denominada ARCHITECT-R. O objetivo do ARCHITECT-R é a geração automática de hardware especializado para robôs móveis através de algoritmos descritos na linguagem CES. A proposta do ARCHITECT-R é bastante arrojada e praticamente impossível de ser desenvolvida por uma só pessoa em um programa de mestrado. Por isso o trabalho que será realizado no decorrer desse mestrado será apenas a criação de uma parte dessa ferramenta. Mais especificamente, serão tratados a criação de blocos arquiteturais (hardwares reconfiguráveis) para redes neurais artificiais utilizados pela linguagem CES e a especificação do sistema.
Title in English
Not available
Keywords in English
Not available
Abstract in English
In the last few years, the interest in robotics (specially in mobile robots) has grown up and many advances have been obtained in this area. But the task of programming a mobile robot still a very difficult task, both in terms of complexity and time spent. Some new languages specially designed for programming robots have been proposed by researchers. The CES language is a example of such a language. It was developed in the Institute of Robotics, CMU-USA, and will be used as the basic robot programming language for this work. Another important, task for the future of mobile robots is the design of dedicated hardware to accelerate the execution of robot's algorithms. This way, the robot will be able to take better and faster decisions. This task involves deep knowledge of the algorithms used and the design of dedicated hardware. An attempt to diminish this difficulties is proposed in this work. This attempt is the development of a reconfigurable robot hardware design tool. This tool is called ARCHITECT-R. The objective of the ARCHITECT-R is the automatic design of dedicated hardware for mobile robots using algorithms' descriptions in the CES language. This is a very ambitious purpose and would be impossible to be achieved by a researcher alone. This way, just a very small part of the tool is treated in this work. Specifically, the generation of dedicated reconfigurable hardware for neural networks and the UML specification of the tool. Many future works, which will have to be done for the complete development of this tool, is proposed in the last chapter of this dissertation.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2018-01-10
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
CeTI-SC/STI
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2024. All rights reserved.