Hybrid Control for Human-Powered Augmentation Exoskeleton

Abstract

Lower-limb exoskeletons have gained considerable interest in recent years for strength augmentation and load carrying. Many methods have been proposed for strength augmentation. Among these approaches, Sensitivity Amplification Control (SAC) algorithm was first proposed in the augmentation applications of Berkeley Lower Extremity Exoskeleton (BLEEX) and achieve significant performance. It just need the information from the exoskeleton robot, thus reducing the complexity of exoskeleton system. However, SAC algorithm is based on the accuracy dynamic models of exoskeleton in different phases which is hard to obtain. This paper presents a novel control strategy: hybrid control. In the proposed hybrid control strategy, swing leg and stance leg are controlled separately. For the control of swing leg, zero force control strategy is employed to let exoskeleton shadow the motion of pilots smoothly. For the control of stance leg, zero load control strategy is utilized to reduce the gravity of load, which lead pilot to move freely and feel more comfortable. We design a exoskeleton system named HUman-powered Augmentation Lower EXoskeleton (HUALEX) and demonstrate the control efficiency of hybrid control in simulation platform. The experimental results show that the proposed hybrid control strategy can reduce interaction force between the pilot and the exoskeleton efficiently.