Controle de vibração torcional em sistemas rotativos usando redes neurais multicamadas / Torsional vibration control in rotating systems using muitilayer neural networks
AUTOR(ES)
Evaldo Khater
DATA DE PUBLICAÇÃO
1998
RESUMO
The purpose of this work is the development of control strategies of torsional vibration in rotating systems, with the objective of minimizing the significant modes of torsional vibration in steadystate. The active control was employed through a multilayer neural network controler, using back-propagation, feeding the system with the same driving electric motor (DC). A experimental model of the rotating system was employed to adjust the theorical model, test the optimal controler (LQR) and emulation the experimental model using a multilayer neural network to train the appropriate controler to the real system. A electronic circuit attached at the end of flexible shaft sends the amplified signal of angular strain measured. Satisfactory results were found both in the computacional simulation and in the experimental tests, showing that a neural controler can be a good choice for real rotating systems
ASSUNTO(S)
redes neurais (computação) neural networks (computer science) torção vibração torsion vibration
ACESSO AO ARTIGO
http://libdigi.unicamp.br/document/?code=vtls000134203Documentos Relacionados
- Automatization and on-line intelligent control of refrigeration systems using artificial neural networks
- Controle tolerante a falhas usando redes neurais adaptativas
- Reconhecimento de faces usando redes neurais e biometria
- Human face recognition using MLP neural networks
- Reconhecimento de faces usando redes neurais e biometria