ARTIFICIAL NEURAL NETWORKS AS A MOTORICAL ASSOCIATIVE MEMORY FOR HUMANOID ROBOT HAND

This paper relates to the artificial neural networks usage for the purpose of associative memory implementation managing humanoid robot hand. Refering to the structures and mechanisms of processing known from neurophysiological studies model of cognitive control was proposed. In the association realization process two different networks were used: a liquid state machine and a one-way associative network similar to BAM.