On the control of lower extremity exoskeleton base on the interaction force of torso

Abstract

This paper presents an ECUST Lower-extremity Exoskeleton (ELE-ROBOT) for augmenting human performance. Most exoskeletons calculate joint torques by using ground reaction forces from foot force sensors. However, this method becomes unreliable in the complex terrain environment. We propose a new control strategy by measuring the human-robot interaction forces using force sensors mounted on the wearer's torso, and placing only switches on the foot. The exoskeleton joint's torques are estimated by this measured human-robot interaction forces. The experiments were conducted to verify the feasibility of this control method, the wearer carried 35kg load, and walked on the treadmill at 1.3m/s with exoskeleton powered on or off, then the data was compared to the wearer's natural walking gait. The results show ELE-ROBOT can reduce a portion of the carried load that is exerted on the wearer significantly and trace the joints motion in real time.