Human like contact forces adjusting for dexterous robot hands

A lot of research have been carried out to improve the performance of multifingered dexterous robot hands, but research on how to allow such dexterous hands to realize human-like performances has not yet been introduced. In this paper, a fuzzy system and neural network-based scheme is proposed to adjust the contact forces between the fingertips and the nut, to control the hand further, allowing it to screw like human hands. The neural network is used as the preprocessor of the fuzzy reference system. Simulations have been performed using a three-fingered hand model to confirm the effectiveness of the proposed scheme