Learning Rules for Dynamic Trajectories in the Cerebellar Cortex Application to a Bipedal Robot

B. Daya (Lebanon), P. Chauvet, A. Garenne (France), and G.A. Chauvet (France and USA)


Coordinated movement, neural network, cerebellum, walking robot, learning rules.


In this paper the problem of motion control of a biped is considered. We develop a new method based on multi network associated with neural biological strategy. This method deals with organization and coordination aspects in an intelligent modeling of human motion. We propose a cooperative hierarchical multi-network model. The learning of the biped is performed using new dynamic learning rules in a recurrent network. Cognitive agents communicate with reactive (non-articulator) agents in order to generate the motion.

Important Links:

Go Back