TORQUE CONTROL WITH RECURRENT NEURAL NETWORKS

Guillaume Jouffroy

2008

Abstract

In the robotics field, a lot of attention is given to the complexity of the mechanics and particularly to the number of degrees of freedom. Also, the oscillatory recurrent neural network architecture is only considered as a black box, which prevents from carefully studying the interesting features of the network’s dynamics. In this paper we describe a generalized teacher forcing algorithm, and we build a default oscillatory recurrent neural network controller for a vehicle of one degree of freedom. We then build a feedback system as a constraint method for the joint. We show that with the default oscillatory controller the vehicle can however behave correctly, even in its transient time from standing to moving, and is robust to the oscillatory controller’s own transient period and its initial conditions. We finally discuss how the default oscillator can be modified, thus reducing the local feedback adaptation amplitude.

References

  1. Buono, P. and M.Golubitsky (2001). Models of central pattern generators for quadruped locomotion. Journal of Mathematical Biology, 42:291-326.
  2. Ghigliazza, R. and P.Holmes (2004). A minimal model of a central pattern generator and motoneurons for insects locomotion. SIAM Journal on Applied Dynamical Systems, 3(4):671-700.
  3. Ijspeert, A. (2001). A connectionnist central pattern generator for the aquatic and terrestrial gaits of a simulated salamander. Biological Cybernetics, 84:331-348.
  4. Ishiguro, A., Otsu, K., Fujii, A., Uchikawa, Y., Aoki, T., and Eggenberger, P. (2000). Evolving and adaptive controller for a legged-robot with dynamicallyrearranging neural networks. In Proceedings of the Sixth International Conference on Simulation of Adaptive Behavior, Cambridge, MA. MIT Press.
  5. Jouffroy, G. and Jouffroy, J. (2006). A simple mechanical system for studying adaptive oscillatory neural networks. IEEE International Conference on Systems, Man and Cybernetics, pages 2584-2589.
  6. Kailath, A. D. . T. (1990). Model-free distributed learning. IEEE Trans. Neural Networks, 1(1):58-70.
  7. Kamimura, A., Kurokawa, H., Yoshida, E., Tomita, K., Murata, S., and Kokaji, S. (2003). Automatic locomotion pattern generation for modular robots. In Proceedings of IEEE International Conference on Robotics and Automation, pages 714-720.
  8. Krishnaprasad, P. and Tsakriris, D. (1995). Oscillations, se(2)-snakes and motion control. New Orleans, Louisiana.
  9. Mori, T., Nakamura, Y., Sato, M., and Ishii, S. (2004). Reinforcement learning for cpg-driven biped robot. Nineteenth National Conference on Artifical Intelligence, pages 623-630.
  10. Taga, G. (1994). Emergence of bipedal locomotion through entrainment among the neuro-musculo-skeletal system and the environment. Physica D, 75:190-208.
  11. Tsung, F. and Cottrell, G. (1993). Phase-space learning for recurrent networks. Technical Report CS93-285, Dept. Computer Science and Engineering, University of California, San Diego.
  12. Weiss, M. (1997). Learning oscillations using adaptive control. International Conference on artifical Neural Networks, pages 331-336.
Download


Paper Citation


in Harvard Style

Jouffroy G. (2008). TORQUE CONTROL WITH RECURRENT NEURAL NETWORKS . In Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO, ISBN 978-989-8111-31-9, pages 109-114. DOI: 10.5220/0001501401090114


in Bibtex Style

@conference{icinco08,
author={Guillaume Jouffroy},
title={TORQUE CONTROL WITH RECURRENT NEURAL NETWORKS},
booktitle={Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,},
year={2008},
pages={109-114},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001501401090114},
isbn={978-989-8111-31-9},
}


in EndNote Style

TY - CONF
JO - Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 2: ICINCO,
TI - TORQUE CONTROL WITH RECURRENT NEURAL NETWORKS
SN - 978-989-8111-31-9
AU - Jouffroy G.
PY - 2008
SP - 109
EP - 114
DO - 10.5220/0001501401090114