Abstract
Wearable power-assisted devices have attracted a lot of attention recently, but technology for improving user comfort remains an important issue. In order to study soft actuators and control systems that adapt to human movement, we have developed a power-assisted glove that supports grasping force. To reduce discomfort when the glove is driven, we applied a control algorithm based on the pattern classification method. The controller detects a holding shape by analyzing the initial movement pattern of the fingers with angle sensors and operates actuators to assist grasping force before completing the grasp movement. This method makes the power-assisted glove operate as if it predicted the intention of the user.
