Compliance Control of Wearable Robotic Fingers for Rehabilitation Applications

Abstract

The rehabilitation quality of stroke patients has drawn much attention in recent years. Researchers have developed various kinds of devices for lower and upper limb rehabilitation functions. In this paper, we present a 3 degreesof- freedom (3-DOF) wearable robotic fingers for rehabilitation of task-oriented training in grasping tasks. A control strategy is proposed for grasping task training by considering user comfort based-on his/her intention as well as safety in task training procedures. The control strategy is divided into two parts: user compliance control and grasping compliance control. In user compliance control, we employed a mass-spring-damper model for grasping operation when the user exerts an intention force. This strategy allows the user to feel a comfortable guidance of the finger movement. In grasping compliance control, a similar model is used when the finger exoskeleton comes into contact with the object, that the system will become aware of the situation immediately and assist an appropriate grasping force for stable grasping. Experimental verification shows that the developed wearable rehabilitation robotic fingers can provide a comfortable fit for the testers and is capable to assist the testers to achieve grasping task training.