Learning Robots to Move: Biological Control Concepts

Kiriazov, Petko
Learning Robots to Move: Biological Control Concepts
Proc. of the 4th CLAWAR Conf. on Climbing and Walking Robots, Sept. 2001, Karlsruhe, Germany, Eds. K. Berns and R. Dillman, ISBN 1 86058 365 2, Professional Engineering Publishing Ltd, pp. 419-426.

Abstract: The basic functionality of mobile robots is to perform point-to-point movements. We are developing a control design methodology based on recently proven biological control concepts. A direct-search approach involving time/energy minimising test control functions is applied when robot movements are unconstrained. The control learning scheme and parameters to be learned are similar to those of the human. For stabilising trajectory-constrained motion, we apply robust decentralised controllers designed using optimal trade-off relations between bounds of model uncertainty and control force limits. The efficiency of the proposed methodology is verified in dynamics-based simulation of biped locomotion.