2023 年 31 巻 2 号 p. 108-113
Aiming at the control of a powered prosthetic hand, this paper reports the development of a tactile sensing system that estimates the motion intention of the prosthetic hand user. Tactile sensors using Polyvinylidene Difluoride (PVDF) film were developed to detect the forearm surface tactile features caused by muscle contraction of the forearm. A neural network was used to classify hand motions using the forearm surface tactile features as inputs. Number of sensors and sensing location on the forearm were discussed in depth in the paper. Experiments showed that by using only 2 tactile sensors placed roughly above the location of the flexor carpi radialis muscle and the extensor carpi radialis longus muscle, the system can classify 6 types of hand motions at an accuracy of about 90% from experiment participants without intensive training.