Development of A Five-finger Prosthetic Hand Using Ultrasonic Motors Controlled by Two EMG Signals
Makoto Ohga*, Mikio Takeda**, Akira Matsuba*, Akira Koike* and Toshio Tsuji***
*Eastern Hiroshima Prefecture Industrial Research Institute
**Hiroshima Prefectural Institute of Science and Technology
***Graduate School of Engineering, Hiroshima University
Received:March 22, 2002Accepted:August 20, 2002Published:December 20, 2002
Keywords:EMG prosthetic hand, five fingers, compliance control, neural network, motion discrimination
This paper describes an EMG controlled prosthetic hand with five fingers actuated by ultrasonic motors. The grasping mechanism proposed is capable of determining the joint angle of each finger in a way that the five fingers can follow the shape of objects and realize a flexible grasp. Estimation of the motion intended by an amputee is realized by using a neural network involving statistical structures from surface myoelectric signals. The present paper shows that discrimination of four or six kinds of limb-functions is realized with a small number of two electrodes by utilizing the frequency characteristics of EMG signals.
Cite this article as:M. Ohga, M. Takeda, A. Matsuba, A. Koike, and T. Tsuji, “Development of A Five-finger Prosthetic Hand Using Ultrasonic Motors Controlled by Two EMG Signals,” J. Robot. Mechatron., Vol.14 No.6, pp. 565-572, 2002.Data files:
Copyright© 2002 by Fuji Technology Press Ltd. and Japan Society of Mechanical Engineers. All right reserved.