Dinâmica e controle de manipuladores robóticos flexíveis, excitados por carregamento não-linear e não-ideal
Carregando...
Arquivos
Data
2017-09-29
Autores
Lima, Jeferson José de [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Manipuladores robóticos leves trazem diversas vantagens quanto ao custo de produção e eficiência na relação carga útil e peso do robô. No entanto a utilização de materiais leves tende a aumentar a flexibilidade estrutural, principalmente na transmissão do torque e nos elos. As propostas de manipuladores apresentadas neste trabalho possuem requisitos conflitantes, pois deseja-se a redução do peso dos mecanismos e aumentar o desempenho, surgem então alguns desafios de engenharia. Desta forma, esta tese trata da modelagem mecânica, análise dinâ- mica e controle híbridos, utilizando atuadores inteligentes, para manipuladores robóticos com características flexíveis. Dois principais modelos de manipuladores robóticos com características flexíveis serão considerados neste trabalho: um para carregamento de carga conectadas a um cabo e outro para manuseio de ferramenta rotativa, ambos com interação não-ideal do elemento manipulado. É possível verificar comportamento caótico ao adicionar a carga aos manipuladores, trazendo assim, maiores desafios a proposta de controle. Na estratégia de controle são considerados como atuadores os motores CC (Corrente Contínua) e atuadores classificados como materiais inteligentes como a Liga de Memória de Forma (LMF), atuando nos elos flexí- veis, e freio Magneto-Reológico (MR-Brake), atuando nas juntas flexíveis. Utilizou-se da rede neural RBFNN para estimativa das variáveis de controle dos atuadores LMF e MR-Brake. O controle DSDRE (Discrete State Dependent Ricatti Equation) é utilizado como a lei de controle. Os resultados numéricos mostram que a adição de atuadores híbridos trazem um ganho considerável no controle de vibração e posicionamento dos manipuladores.
Light-weight robotic manipulators have advantages over the cost of production and efficiency in relation to payload and weight of the robot. However, the use of light-weight materials tends to increase structural flexibility in torque transmission and in the links of the robot. The manipulator proposals presented in this work have conflicting requirements since it is desired to reduce the weight of the mechanisms and increase the performance, it makes some engineering challenges. Thus, this thesis deals with mechanical modeling, dynamic analysis and hybrid control, using intelligent actuators, for robotic manipulators with flexible characteristics. Two models of robotic manipulators with flexible characteristics will be considered in this work: one for payload connected to a cable and another one for rotary tool, both with non-ideal interaction of the manipulated element. It is possible to verify chaotic behavior by adding the load to the manipulators, becoming a great challenger to the proposal control. DC motors and actuators classified as intelligent materials such as the shape memory alloy (SMA), acting on the flexible links, and Magneto-Rheological Brake (MR-Brake), acting on the flexible joints, are used in the strategy control. The neural network RBF was used to estimate the control variables of the SMA and MR-Brake actuators. The DSDRE (Discrete State Dependent Ricatti Equation) control is used as the control law. The numerical results show that the addition of hybrid actuators possibly have a considerable gain in the control of vibration and positioning of the manipulators.
Light-weight robotic manipulators have advantages over the cost of production and efficiency in relation to payload and weight of the robot. However, the use of light-weight materials tends to increase structural flexibility in torque transmission and in the links of the robot. The manipulator proposals presented in this work have conflicting requirements since it is desired to reduce the weight of the mechanisms and increase the performance, it makes some engineering challenges. Thus, this thesis deals with mechanical modeling, dynamic analysis and hybrid control, using intelligent actuators, for robotic manipulators with flexible characteristics. Two models of robotic manipulators with flexible characteristics will be considered in this work: one for payload connected to a cable and another one for rotary tool, both with non-ideal interaction of the manipulated element. It is possible to verify chaotic behavior by adding the load to the manipulators, becoming a great challenger to the proposal control. DC motors and actuators classified as intelligent materials such as the shape memory alloy (SMA), acting on the flexible links, and Magneto-Rheological Brake (MR-Brake), acting on the flexible joints, are used in the strategy control. The neural network RBF was used to estimate the control variables of the SMA and MR-Brake actuators. The DSDRE (Discrete State Dependent Ricatti Equation) control is used as the control law. The numerical results show that the addition of hybrid actuators possibly have a considerable gain in the control of vibration and positioning of the manipulators.
Descrição
Palavras-chave
Manipulador robótico flexíveis, Sistemas não-ideias, Materiais Inteligentes, Controle DSDRE, Flexible robotic manipulator, Non-ideal systems, Intelligent materials, DSDRE control