Neural Computing and Applications

, Volume 19, Issue 1, pp 103–114 | Cite as

A neural framework for adaptive robot control

Original Article

Abstract

This paper investigates how dynamics in recurrent neural networks can be used to solve some specific mobile robot problems such as motion control and behavior generation. We have designed an adaptive motion control approach based on a novel recurrent neural network, called Echo state networks. The advantage is that no knowledge about the dynamic model is required, and no synaptic weight changing is needed in presence of time varying parameters in the robot. To generate the robot behavior over time, we adopted a biologically inspired approach called neural fields. Due to its dynamical properties, a neural field produces only one localized peak that indicates the optimum movement direction, which navigates a mobile robot to its goal in an unknown environment without any collisions with static or moving obstacles.

Keywords

Recurrent neural networks Metalearning Mobile robots Navigation Adaptive control 

References

  1. 1.
    Kolmanovsky I, McClamroch H (1995) Development in nonholonomic control problems. IEEE Control Systems 20–36Google Scholar
  2. 2.
    Fierro R, Lewis F (1998) Control of a nonholonomic mobile robot using neural networks. IEEE Trans Neural Netw 9(4):589–600CrossRefGoogle Scholar
  3. 3.
    Fukao T, Nakagawa H, Adachi N (2000) Adaptive tracking control of a nonholonomic mobile robot. IEEE Trans Robot Autom 16(5):609–615CrossRefGoogle Scholar
  4. 4.
    Khatib O (1986) Real-time obstacle avoidance for manipulators and mobile robots. Int J Robotic Res 5(1):90–98CrossRefMathSciNetGoogle Scholar
  5. 5.
    Koren Y, Borenstein J (1991) Potential field methods and their inherent limitations for mobile robot navigation. In: Proceedings of the IEEE ICRA, Sacramento, California, pp 1398–1404Google Scholar
  6. 6.
    Amari S (1977) Dynamics of pattern formation in lateral-inhibition type neural fields. Biol Cybern 27:77–87MATHCrossRefMathSciNetGoogle Scholar
  7. 7.
    Kechriotis G, Zervas E, Manolakos E (1994) Using recurrent neural networks for adaptive communication channel equalization. IEEE Trans Neural Netw 5:267–278CrossRefGoogle Scholar
  8. 8.
    Chao-Chee K, Lee KY (1995) Diagonal recurrent neural networks for dynamic systems control. IEEE Trans Neural Netw 6:144–156Google Scholar
  9. 9.
    Quanmin Z, Guo L (2004) Stable adaptive neurocontrol for nonlinear discrete-time systems. IEEE Trans Neural Netw 15:653–662Google Scholar
  10. 10.
    Connor JT, Martin RD, Atlas LE (1994) Recurrent neural networks and robust time series prediction. IEEE Trans Neural Netw 5:240–254CrossRefGoogle Scholar
  11. 11.
    Harter D, Kozma R (2005) Chaotic neurodynamics for autonomous agents. IEEE Trans Neural Netw 16:565–579CrossRefGoogle Scholar
  12. 12.
    Williams RJ, Zipser D (1989) A learning algorithm for continually running fully recurrent neural networks. Neural Comput 1:270–280CrossRefGoogle Scholar
  13. 13.
    Perez-Ortiz JA, Gers FA, Eck D, Schmidhuber J (2003) Kalman filters improve lstm network performance in problems unsolvable by traditional recurrent nets. Neural Netw 2:241–250CrossRefGoogle Scholar
  14. 14.
    Juergen Schmidhuber (1992) A fixed size storage o(n3) time complexity learning algorithm for fully recurrent continually running networks. Neural Comput 4(2):243–248Google Scholar
  15. 15.
    Werbos PJ (1990) Backpropagation through time: what it does and how to do it. In: Proceedings of the IEEE, vol 78, pp 1550–1560Google Scholar
  16. 16.
    Jaeger H (2002) Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the echo state network approach. Technical Report 159, AIS Fraunhofer, St. Augustin, GermanyGoogle Scholar
  17. 17.
    Jaeger H (2001) The ’echo state’ approach to analysing and training recurrent neural networks. Technical Report 148, AIS Fraunhofer, St. Augustin, GermanyGoogle Scholar
  18. 18.
    Wenjie D, Huo W, Tso SK, Xu WL (2000) Tracking control of uncertain dynamic nonholonomic system and its application to wheeled mobile robots. IEEE Trans Robot Autom 16(6)Google Scholar
  19. 19.
    Oubbati M, Schanz M, Levi P (2005) Kinematic and dynamic adaptive control of a nonholonomic mobile robot using a RNN. In: Proceedings of the 6th IEEE symposium on computational intelligence in robotics and automation, pp 27–33Google Scholar
  20. 20.
    Oubbati M, Schanz M, Buchheim T, Levi P (2005) Velocity control of an omnidirectional robocup player with recurrent neural networks. In: RoboCup, Osaka, Japan, pp 691–701Google Scholar
  21. 21.
    Vilalta R, Drissi Y (2002) A perspective view and survey of meta-learning. Artif Intell Rev 18:77–95CrossRefGoogle Scholar
  22. 22.
    Cotter NE, Conwell PR (1990) Fixed-weight networks can learn. In: International joint conference on neural networks, vol 3, pp 553–559Google Scholar
  23. 23.
    Feldkamp LA, Puskorius GV, Moore PC (1996) Adaptation from fixed weight dynamic networks. In: Proceedings IEEE international conference on neural networks, pp 155–160Google Scholar
  24. 24.
    Feldkamp LA, Puskorius GV, Moore PC (1997) Fixed weight Controller for multiple systems. In: Proceedings of IEEE international conference on neural networks, pp 773–778Google Scholar
  25. 25.
    Steven Younger A, Conwell PR, Cotter NE (1999) Fixed weight on line learning. IEEE Trans Neural Netw 10(2):272–283CrossRefGoogle Scholar
  26. 26.
    Prokhorov D, Puskorius G, Feldkamp L (2001) Dynamical neural networks for control. IEEE PressGoogle Scholar
  27. 27.
    Oubbati M, Schanz M, Levi P (2005) Meta-learning for adaptive identification of non-linear dynamical systems. In: Proceedings of 20th international symposium on intelligent control, IEEE, pp 473–478Google Scholar
  28. 28.
    Oubbati M, Schanz M, Levi P (2005) A fixed-weight RNN dynamic controller for multiple mobile robots. In: Proceedings of 24th IASTED international conference on modelling, identification, and control, Innsbruck, Austria, pp 277–282Google Scholar
  29. 29.
    Schöner G, Dose M, Engels C (1995) Dynamics of behavior: theory and applications for autonomous robot architectures. Robot Auton Syst 16Google Scholar
  30. 30.
    Steinhage A, Schöner G (1997) Self-calibration based on invariant view recognition: dynamic approach to navigation. Robot Auton Syst 20:133–156CrossRefGoogle Scholar
  31. 31.
    Steinhage A, Schöner G (1997) The dynamic approach to autonomous robot navigation. In: ISIE97, IEEE international symposium on industrial electronicsGoogle Scholar
  32. 32.
    Bergener T, Bruckhoff C, Dahm P, Janßen H, Joublin F, Menzner R, Steinhage A, von Seelen W (1999) Complex behavior by means of dynamical systems for an anthropomorphic robot. Neural Netw 12(7–8):1087–1099CrossRefGoogle Scholar
  33. 33.
    Goldenstein S, Metaxis DM, Large EW (2000) Nonlinear dynamic systems for autonomous agent navigation. In: Proceedings of the 17th national conference on artificial intelligenceGoogle Scholar
  34. 34.
    Monteiro S, Bicho E (2002) A dynamical systems approach to behavior-based formation control. In: Proceedings of the 2002 IEEE ICRAGoogle Scholar
  35. 35.
    Dahm P, Bruckhoff C, Joublin F (1998) A neural field approach to robot motion control. In: Proceedings of the IEEE international conference on systems, man, and cybernetics, pp 3460–3465Google Scholar
  36. 36.
    Giese MA (2000) Neural field model for the recognition of biological motion. In: Second international ICSC symposium on neural computation, Berlin, GermanyGoogle Scholar
  37. 37.
    Edelbrunner H, Handmann U, Igel C, Leefken I, von Seelen W (2001) Application and optimization of neural field dynamics for driver assistance. In: IEEE 4th international conference on intelligent transportation systems, IEEE Press, pp 309–314Google Scholar
  38. 38.
    Erlhagen W, Bicho E (2006) The dynamic neural field approach to cognitive robotics. J Neural Eng 3:R36–R54Google Scholar
  39. 39.
    Oubbati M, Schanz M, Levi P (2006) Neural fields for behaviour-based control of mobile robots. In: 8th international IFAC symposium on robot control. Bologna, ItalyGoogle Scholar
  40. 40.
    Oubbati M, Palm G (2007) Neural fields for controlling formation of multiple robots. In: IEEE international symposium on computational intelligence in robotics and automation, pp 90–94Google Scholar
  41. 41.
    Kishimoto K, Amari S (1979) Existence and stability of local excitations in homogeneous neural fields. J Math Biol 7:303–318MATHCrossRefMathSciNetGoogle Scholar
  42. 42.
    Kalmár-Nagy T, D’Andrea R, Ganguly P (2004) Near-optimal dynamic trajectory generation and control of an omnidirectional vehicle. Robot Auton Syst 46:47–64Google Scholar

Copyright information

© Springer-Verlag London Limited 2009

Authors and Affiliations

  1. 1.Institute of Neural Information ProcessingUniversity of UlmUlmGermany

Personalised recommendations