A Study to Improve Operating Accuracy of Electric Prosthetic Hand Using Tactile Feature Sensing System

Abstract

　This paper reports on a performance study of an electric prosthetic hand using a new electric prosthetic hand control method, the tactile feature sensing system. This prosthetic hand is controlled by attaching tactile sensors made of polyvinylidene fluoride (PVDF) to the forearm skin surface to detect changes in tactile properties such as forearm muscle protuberance, contraction, and stiffening. This paper investigates the influence of fatigue, perspiration, and other conditions that may occur after prolonged use of a prosthetic hand, on the accuracy of hand motion classification. The results showed that perspiration did not deteriorate the accuracy of hand motion classification but serve fatigue conditions reduced accuracy. Therefore, to solve this problem, we attempted to improve accuracy by changing or adding feature values to be evaluated.