Hierarchically intelligent control of a bionic arm

Abstract

A shoulder-disarticulation prosthesis with seven degrees of freedom is formulated as a decision process accepting fuzzy commands from the human nervous system. The problem is decomposed into a functional hierarchy of three levels associated with organization, coordination, and Self-Organizing control, respectively. The highest level transforms a complex command into a sequence of elementary motions. A fuzzy automaton in the middle level coordinates the action of the lowest level controllers which apply direct control inputs to the nonlinear plant.