S22  Mecatrônica    Titulo: ARTIFICIAL NEURAL NETWORKS TO KINEMATICS MODELING OF ROBOT MANIPULATORS   Resumo : ONE OF THE PROBLEMS IN ROBOTICS, NOWADAYS, IS THE ACQUISITION OF MATHEMATICAL MODELS OF ROBOTIC MANIPULATORS. IT IS DUE, MAINLY, TO THE COMPLEXITY OF THE MANIPULATOR, THAT MAKES DIFFICULTY THE WAY TO OBTAIN OF SOME PARAMETERS RELATED WITH THE GEOMETRICAL AND DYNAMICAL MODELS. THIS WORK PRESENTS NA ALTERNATIVE APPROACH TO DECREASE THESE PROBLEMS, CONSISTING IN THE USE OF ARTIFICIAL NEURAL NETWORKS TO REPRESENT THE FORWARD AND INVERSE KINEMATICS MODELS OF A ROBOTIC MANIPULATOR WITH 3 DEGREES OF FREEDOM. THE RESULTS SHOWS THIS IS A GOOD TECHNIQUE TO REPRESENT THOSE MODELS. FURTHERMORE, THE QUALITY OF RESULTS DEPENDS ON THE STRUCTURE OF THE NEURAL NETWORK AND ON THE LEARNING STRATEGY ALSO. KEYWORDS: ROBOTIC MANIPULATOR, KINEMATICS MODEL, ARTIFICIAL NEURAL NETWORKS     Autores : de Oliveira, Vinicius Menezes Gomes, Sebastião C. P.     Trabalho Completo :   COBEM99 | Comissão  Organizadora | Palestras | Sessões Técnicas | Autor | Simpósios e Sessões Especiais | Revisores |  Título  dos  Trabalhos | Local