Anais Do 5. Congresso Brasileiro De Redes Neurais 2016
DOI: 10.21528/cbrn2001-101
|View full text |Cite
|
Sign up to set email alerts
|

Identifica��o e Compensa��o de Atritos N�o Lineares de Atuadores Rob�ticos Via Redes Neurais Artificiais

Abstract: IntroduçãoExistem atualmente diversas aplicações de redes neurais artificiais (RNA), nos mais variados domínios da ciência e tecnologia [(Kaynak, and Ertugru, 1997), (Jung and Hsia, 1998)]. Trata-se de um assunto que tem merecido grande atenção por parte da comunidade científica. Em [Miller, 1995] tem uma importante descrição sobre o histórico das redes neurais artificiais.Neste artigo investiga-se a possibilidade de identificação do torque de atrito de um atuador robótico do tipo moto-redutor, utilizando-se u… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
2
0
1

Year Published

2016
2016
2020
2020

Publication Types

Select...
2

Relationship

1
1

Authors

Journals

citations
Cited by 2 publications
(3 citation statements)
references
References 2 publications
0
2
0
1
Order By: Relevance
“…Its also represents the input values (X 1 , X 2 , ..., X 10 ) used by the neurons to calculate the outputs (Y 1 , Y 2 ) of the ANN. Among the literature found on the use of compensation methods with artificial neural networks, the studies Jang and Jeon (2000), Gervini and Gomes (2001), Gervini et al (2003) and Selmic and Lewis (2000) stand out. These researches presented different ways of using neural networks to compensate the dead zone and are the studies most cited by publications on the same topic.…”
Section: Artificial Neural Networkmentioning
confidence: 99%
See 1 more Smart Citation
“…Its also represents the input values (X 1 , X 2 , ..., X 10 ) used by the neurons to calculate the outputs (Y 1 , Y 2 ) of the ANN. Among the literature found on the use of compensation methods with artificial neural networks, the studies Jang and Jeon (2000), Gervini and Gomes (2001), Gervini et al (2003) and Selmic and Lewis (2000) stand out. These researches presented different ways of using neural networks to compensate the dead zone and are the studies most cited by publications on the same topic.…”
Section: Artificial Neural Networkmentioning
confidence: 99%
“…A sequence of papers published by S. C. P. Gomes and V. I. Gervini draws attention. Initially, in the article Gervini and Gomes (2001) a training strategy is presented and a structure of a neural network is proposed to learn the torque friction of a robotic actuator. The authors created and trained a backpropagation neural network with two input neurons to receive the motor torque and rotor speed variables and estimate an output referring to the estimated friction torque variable, this being the value responsible for compensating the losses from the dead zone.…”
Section: Artificial Neural Networkmentioning
confidence: 99%
“…Utilizou-se a arquitetura de rede proposta em [10], a qual é composta de dois neurônios na camada de entrada (torque motor e velocidade do rotor), quatro na camada intermediária e um na camada de saída (torque de compensação neural), a qual é mostrada na fig. 4.…”
Section: Controleunclassified