Force/Position Hybrld Control for Geometrlcally Constrained Tendon-Driven Flexible Joint Arm

The purpose of this study was to design force/position hybrid controllers for a tendon-driven flexible joint arm. Tendon-driven flexible joint arms are expected to be more suitable for force control tasks than the usual motor-driven robot arms because of their weightlessness and their stability for contact. Combining the joint-space orthogonalization method (Liu et al., (1998)) and the tendon-driven robotic mechanism controller (Kobayashi et al., (2003)), we construct two types of controllers, a non-adaptive controller and a neural networks(NN) controller, and show their stability. Simulation shows them to be efficient.