Interaction control of a programmable footpad-type gait rehabilitation robot for active walking on various terrains

Abstract

This paper discusses interaction control of a programmable footpad-type gait rehabilitation robot which allows walking velocity updates on various terrains for patients' active walking. In human, there exists a neuronal coordination of arm and leg. When humans walk, both the upper and lower limb are moving together synchronized at one to one frequency. Additionally it has been known that providing arm swing could be beneficial during locomotor training. Thus, the suggested rehabilitation robot is composed of an upper limb device, a sliding device, two footpad devices, and a partial body support system. The footpad device on the slide device can generate various trajectories such as planar, stairs and slope walking. The upper-limb device allows users to swing their arms naturally through the use of a simple pendulum link with a passive prismatic joint. The upper and lower limbs of the robot are connected by a synchronized gait pattern. By analyzing the interaction torque between the upper limb of the human and the robot, the walking velocity on various terrains can be updated. Experimental results with a healthy subject show that the user can update the walking velocity on level ground, slopes, and stairs through upper and lower limb connections. Hence, the training might give more effective result since the robot allows the patient to walk according to his intentions on various terrains.