A neural network classifer for robot hand preshape and grasp task matching

Bu makalede, robot eli tutuş biçimi ve tutuş görevi eşleştirmesi için tasarlanan yapay sinir ağı sınıflandırıcı sunulmaktadır. Robot eli tutuş biçimi, sade bir gösterimle, hassas tutuşlar için, tutuşa dahil olan parmak sayısı ve robot elin nesneye yaklaşma yönü ile ifade edilmiştir. Sınıflandırıcı, tutuş görevi ve nesne özelliklerini girdi olarak almakta ; çıktı olarak ise tutuş biçimini vermektedir. Sınıflandırıcı, başlangıçta eğitilmektedir. Tasarlanan sınıflandırıcı robot eli tutuş sınamasında denenmiştir.

Robot el tutuş biçimi ve tutuş görevi eşleştirmesi için yapay sinir ağı sınıflandırıcı

In this paper, a neural network classifier for grasping task and robot hand preshape matching is presented, which is based on a plain representation of robot hand grasp preshape for precision grasps, described by number of fingers to be involved in grasp and approach direction of hand towards object. The classifier is initially trained and takes the desired grasp task and object attributes as input and yields the preshape as output. The classifier is tested by using robot hand grasp preshape simulation .

___

1. Bicchi,A. and Kumar,V., “Robotic Grasping and Contact: A Review”, In Proc. IEEE ICRA 2000, San Francisco, CA., 2000.
2. Okamura,A.M. et al., “An overview of dexterousmanipulation.” IEEE International Conference on Robotics and Automation, San Francisco, CA. 255-262, 2000.
3. Shimoga,K.B.,”Robot grasp synthesis algorithms: A survey”, International Journal of Robotics Research, 15 (3):230-266, June 1996.
4. Cutkosky,M.R., “ On grasp choice grasp models and the design of hands for manufacturing tasks ”, IEEE Trans. Robotics and Automation, Vol. 5, No. 3., 1989.
5. Iberall,T., “The Nature of Human Prehension: Three Dextrous Hands in One,'' Proc. of the IEEE Int. Conf.on Robotics and Automation, pp. 396-401, 1987.
6. Lyons,D.M , “ A simple set of grasps for a dextrous hand ”, Proc. Of IEEE Int. Conf. Robotics & Automation, pp.588-594, 1985.
7. Baysal,C.V., “Preshaping Control for Dexterous Manipulation with a Multifingered Robot Hand”, PhD Thesis, Dept.of Electrical and Electronics Engineering, Cukurova University, 2006.
8. Haykin,S., “Neural Networks-A Comprehensive Foundation”, Macmiilan College Publishing, 1994.
9. Pollard,N. S., “Closure and Quality Equivalence for Efficient Synthesis of Grasp from Examples”, The International Journal of Robotics Research, Vol 23, No 6, pp595-613, June 2004.