Home > Online first >

Implementation of Fall Detection for WAR-Exoskeleton Robot

Abstract—In this paper, we presented a method of detecting the state of the lower limb exoskeleton for a fall detection function based on the WAR exoskeleton robot platform. While in motion, the exoskeleton is to follow the gait pattern data. The exoskeleton functions on distributive control where the hip joints and knee joints are controlled by the PID controller. Several equations are derived which outlines the conditions needed to keep the exoskeleton balanced and unlikely to fall. Deviations from the expected patterns and angle changes in the exoskeleton robot may cause instability and imbalance. Experimental results on the moment during leg motion and positional changes of the centroid while robot is in motion are obtained and used as conditions for fall detection. The method proposed is effective for fall detection in war-exoskeleton robot.  
Index Terms—fall detection; assistive device; exoskeleton robot; gait pattern; lower limb motion