Fuzzy-based grasp-force-adaptation for multifingered robot hands

A fuzzy logic approach for the online grasp-force-adaptation, which can be used for the control of fine manipulating with a multifingered robot hand, will be presented in this paper. The kernel of this approach consists of decision making logic which expresses the a-priory knowledge about the force behaviour inside the friction cones and the necessary reactions, regarding the criterion for grip stability. For the software realisation of the fuzzy control approach a Fuzzy-C Development System supporting the entire development process was used. Two corresponding fuzzy controllers have been designed: A finger controller and a grasp controller. During fine manipulations of an object the fuzzy controller interacts with an underlying conventional controller that receives, after defuzzyfication, the adapted force values which were applied. A computer based simulation system was developed to analyse the capabilities of the designed fuzzy controllers.<<ETX>>