[17a-K306-1~10] FS.1 Focused Session "AI Electronics"

9:30 AM - 9:45 AM JST(12:30 AM - 12:45 AM UTC)

[17a-K306-3] Improvement of Controllability of Myoelectric Prosthesis Hand Based on Reservoir Computing Framework

〇Yusuke Hoshika¹, Seiya Kasai¹ (1.RCIQE, Hokkaido Univ.)

Keywords:

Myoelectric prosthesis hand,Surface myoelectric signal analysis,Reservoir computing

A big issue in a myoelectric prosthetic hand is poor controllability, which is attributed to the difficulty in readout of the intended motion from weak and complicated surface myoelectric signals. Here, we consider that the complicated myoelectric signals correspond to the output signal from the reservoir layer in a reservoir computing system. Accordingly, we developed a setup which learns and reproduces the intended motion by linear combination of the myoelectric signals induced from multiple electrode positions. Using the learned model for five movements, we attempted real-time control of a robot arm. As a result, three out of the five movements were successfully reproduced.