design of prosthetic robot hand and electromyography-based

7
ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37 ๊ถŒ์ œ 5 ํ˜ธ pp. 339-345 May 2020 / 339 J. Korean Soc. Precis. Eng., Vol. 37, No. 5, pp. 339-345 http://doi.org/10.7736/JKSPE.019.138 ISSN 1225-9071 (Print) / 2287-8769 (Online) ๋กœ๋ด‡ ์˜์ˆ˜ ์„ค๊ณ„ ๋ฐ ๊ทผ์ „๋„ ๊ธฐ๋ฐ˜์˜ ์†๋™์ž‘ ์ธ์‹ Design of Prosthetic Robot Hand and Electromyography-Based Hand Motion Recognition ์žฅํ˜ธ๋ช… 1 , ์†์ •์šฐ 1,# Ho Myoung Jang 1 and Jung Woo Sohn 1,# 1 ๊ธˆ์˜ค๊ณต๊ณผ๋Œ€ํ•™๊ต ๊ธฐ๊ณ„์„ค๊ณ„๊ณตํ•™๊ณผ (Department of Mechanical Design Engineering, Kumoh National Institute of Technology) # Corresponding Author / E-mail: [email protected], TEL: +82-54-478-7378 ORCID: 0000-0003-3962-7380 KEYWORDS: Prosthetic robot hand ( ๋กœ๋ด‡ ์˜์ˆ˜), Surface electromyography ( ํ‘œ๋ฉด ๊ทผ์ „๋„), Hand motion recognition ( ์†๋™์ž‘ ์ธ์‹), Artificial neural network ( ์ธ๊ณต ์‹ ๊ฒฝ๋ง) In this paper, a prosthetic robot hand was designed and fabricated and experimental evaluation of the realization of basic gripping motions was performed. As a first step, a robot finger was designed with same structural configuration of the human hand and the movement of the finger was evaluated via kinematic analysis. Electromyogram (EMG) signals for hand motions were measured using commercial wearable EMG sensors and classification of hand motions was achieved by applying the artificial neural network (ANN) algorithm. After training and testing for three kinds of gripping motions via ANN, it was observed that high classification accuracy can be obtained. A prototype of the proposed robot hand is manufactured through 3D printing and servomotors are included for position control of fingers. It was demonstrated that effective realization of gripping motions of the proposed prosthetic robot hand can be achieved by using EMG measurement and machine learning-based classification under a real-time environment. Manuscript received: October 15, 2019 / Revised: January 29, 2020 / Accepted: February 10, 2020 This paper was presented at KSPE Spring Conference 2019 1. ์„œ๋ก  ์˜์ˆ˜๋Š” ์ƒ์ง€ ์ ˆ๋‹จ์ž์˜ ์ผ์ƒ์ƒํ™œ์„ ๋•๊ธฐ ์œ„ํ•œ ๋ณด์กฐ ๊ธฐ๊ตฌ๋กœ ํฌ ๊ฒŒ ์žฅ์‹ ์˜์ˆ˜( ๋ฏธ์šฉ ์˜์ˆ˜), ์ž‘์—… ์˜์ˆ˜, ์ธ์ฒด ๊ตฌ๋™ํ˜• ์˜์ˆ˜, ์ „๋™ ์˜ ์ˆ˜ ๋“ฑ์œผ๋กœ ๋ถ„๋ฅ˜ํ•  ์ˆ˜ ์žˆ๋‹ค. ์žฅ์‹ ์˜์ˆ˜๋Š” ์™ธํ˜•๋งŒ ์†์˜ ๋ชจ์–‘๊ณผ ๋น„ ์Šทํ•˜๊ฒŒ ํ•œ ๊ฒƒ์œผ๋กœ ๋™์ž‘์€ ๋ถˆ๊ฐ€๋Šฅํ•œ ๊ฒƒ์ด๊ณ , ์ž‘์—… ์˜์ˆ˜๋Š” ๊ฐˆ๊ณ ๋ฆฌ ๋ชจ์–‘์˜ ์†๋์œผ๋กœ ๋ฌผ๊ฑด์„ ์žก๊ฑฐ๋‚˜ ์ด๋™์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š” ํ˜•ํƒœ์ด๋‹ค. ์ธ ์ฒด ๊ตฌ๋™ํ˜• ์˜์ˆ˜๋Š” ์–ด๊นจ ์›€์ง์ž„ ๋“ฑ์„ ์ด์šฉํ•ด์„œ ๊ฐ„๋‹จํ•œ ๋™์ž‘์„ ๊ตฌ ํ˜„ํ•  ์ˆ˜ ์žˆ์œผ๋‚˜, ์ธ์ฒด ํŠน์ • ๋ถ€๋ถ„์— ๋ฌด๋ฆฌ๋ฅผ ์ค„ ์ˆ˜ ์žˆ๋Š” ๋‹จ์ ์ด ์žˆ ๋‹ค. ์ „๋™ ์˜์ˆ˜๋Š” ๋ชจํ„ฐ๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๊ฐ€๋ฝ์„ ์›€์ง์ผ ์ˆ˜ ์žˆ๋Š” ํ˜• ํƒœ๋กœ ์ œํ•œ๋œ ์†๋™์ž‘์„ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ์ง€๋งŒ, ๊ฐ€๊ฒฉ์ด ๋งค์šฐ ๋†’์•„ ์ œ ํ•œ์ ์œผ๋กœ ํ™œ์šฉ๋˜๊ณ  ์žˆ๋‹ค. ์ตœ๊ทผ์—๋Š” ์‹ ์ฒด์˜ ์ž”์กด ๊ทผ์œก์—์„œ ์ธก์ • ํ•  ์ˆ˜ ์žˆ๋Š” ํ‘œ๋ฉด ๊ทผ์ „๋„๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๋™์ž‘์„ ๊ตฌํ˜„ํ•˜๋Š” ๋‹ค์–‘ํ•œ ํ˜•ํƒœ์˜ ์˜์ˆ˜๋“ค์ด ์ œ์•ˆ๋˜๊ณ  ์žˆ์œผ๋ฉฐ ๊ด€๋ จ ์—ฐ๊ตฌ๊ฐ€ ํ™œ๋ฐœํ•˜๊ฒŒ ์ง„ํ–‰ ๋˜๊ณ  ์žˆ๋‹ค. Saridis ์™€ Gootee 1 ๋Š” ์ด๋‘๊ทผ๊ณผ ์‚ผ๋‘๊ทผ์˜ ๊ทผ์ „๋„๋ฅผ ์ธก์ •ํ•˜์—ฌ ์ „์™„์˜ ๋™์ž‘์„ ๊ตฌํ˜„ํ•˜๋Š” ์˜์ˆ˜๋ฅผ ์ œ์•ˆํ•˜๊ณ  ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ๊ฒ€์ฆ ํ•˜์˜€๋‹ค. Zero Crossing ๊ณผ ๋ถ„์‚ฐ์„ ์ด์šฉํ•˜์—ฌ ํŠน์ง•์„ ์ถ”์ถœํ•˜๊ณ , ๊ณ„ ์ธต์  ์ง€๋Šฅ ์ œ์–ด๊ธฐ๋ฅผ ์ด์šฉํ•˜์—ฌ ํŒจํ„ด์„ ๋ถ„๋ฅ˜ํ•˜์˜€๋‹ค. Matrone 2 ๋“ฑ ์€ ์ฃผ์„ฑ๋ถ„ ๋ถ„์„ ๊ธฐ๋ฐ˜์˜ ์ œ์–ด๊ธฐ๋ฅผ ์ด์šฉํ•œ ๊ณผ์†Œ ๊ตฌ๋™ ์ „๋™ ์˜์ˆ˜์˜ ์ œ์–ด๋ฅผ ์ œ์•ˆํ•˜๊ณ , ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํšจ๊ณผ๋ฅผ ๊ณ ์ฐฐํ•˜์˜€๋‹ค. Ma 3 ๋“ฑ์€ ๊ทผ์œก ์‹œ๋„ˆ์ง€ ๋ชจ๋ธ์„ ์ด์šฉํ•˜์—ฌ ๊ทผ์ „๋„ ๊ธฐ๋ฐ˜ ์˜์ˆ˜์˜ ์†๊ณผ ์†๋ชฉ ์šด ๋™์„ ์ œ์–ดํ•˜๋Š” ๊ฒƒ์„ ์ œ์•ˆํ•˜์˜€๊ณ , ๊ฐ€์ƒ ๊ณต๊ฐ„์—์„œ ์˜์ˆ˜๊ฐ€ ๋™์ž‘์„ ํšจ๊ณผ์ ์œผ๋กœ ๊ตฌํ˜„ํ•˜๋Š” ๊ฒƒ์„ ํ™•์ธํ•˜์˜€๋‹ค. Guo 4 ๋“ฑ์€ ๊ทผ์ „๋„ ์‹ ํ˜ธ ์™€ ํ•จ๊ป˜ ๊ทผ์ ์™ธ์„  ๋ถ„๊ด‘๋ฒ•(Near-Infrared Spectroscopy, NIRS) ์„ ํ•จ๊ป˜ ์ด์šฉํ•˜์—ฌ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋งŒ์„ ์ด์šฉํ•  ๋•Œ๋ณด๋‹ค ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋†’์ผ ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•˜๊ณ  ๊ฐ€์ƒ ๊ณต๊ฐ„์˜ ์˜์ˆ˜๋ฅผ ์ด์šฉํ•˜์—ฌ ๊ฒ€ ์ฆํ•˜์˜€๋‹ค. Lee 5,6 ๋“ฑ์€ ๊ทผ์ „๋„ ์ธก์ •์„ ์œ„ํ•œ ์ „๊ทน์ด ํฌํ•จ๋˜์–ด ์žˆ ๋Š” ๋‹ˆํŠธ ๋ฐด๋“œ ํ˜•ํƒœ์˜ ์žฅ์น˜๋ฅผ ์ œ์•ˆํ•˜๊ณ , ์‹คํ—˜์œผ๋กœ ํšจ์œจ์„ฑ์„ ๊ฒ€์ฆ ํ•˜์˜€๋‹ค. Lee 7 ๋“ฑ์€ ์˜์ˆ˜์˜ ํŒŒ์ง€ ์„ฑ๋Šฅ ํ‰๊ฐ€ ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•˜๊ณ , ์ด๋ฅผ Copyright ยฉ The Korean Society for Precision Engineering This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/ 3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Upload: others

Post on 16-Apr-2022

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Design of Prosthetic Robot Hand and Electromyography-Based

ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ pp. 339-345 May 2020 / 339

J. Korean Soc. Precis. Eng., Vol. 37, No. 5, pp. 339-345 http://doi.org/10.7736/JKSPE.019.138

ISSN 1225-9071 (Print) / 2287-8769 (Online)

๋กœ๋ด‡ ์˜์ˆ˜ ์„ค๊ณ„ ๋ฐ ๊ทผ์ „๋„ ๊ธฐ๋ฐ˜์˜ ์†๋™์ž‘ ์ธ์‹

Design of Prosthetic Robot Hand and Electromyography-Based HandMotion Recognition

์žฅํ˜ธ๋ช…1, ์†์ •์šฐ1,#

Ho Myoung Jang1 and Jung Woo Sohn1,#

1 ๊ธˆ์˜ค๊ณต๊ณผ๋Œ€ํ•™๊ต ๊ธฐ๊ณ„์„ค๊ณ„๊ณตํ•™๊ณผ (Department of Mechanical Design Engineering, Kumoh National Institute of Technology)

# Corresponding Author / E-mail: [email protected], TEL: +82-54-478-7378

ORCID: 0000-0003-3962-7380

KEYWORDS: Prosthetic robot hand (๋กœ๋ด‡ ์˜์ˆ˜), Surface electromyography (ํ‘œ๋ฉด ๊ทผ์ „๋„), Hand motion recognition (์†๋™์ž‘ ์ธ์‹),

Artificial neural network (์ธ๊ณต ์‹ ๊ฒฝ๋ง)

In this paper, a prosthetic robot hand was designed and fabricated and experimental evaluation of the realization of basic

gripping motions was performed. As a first step, a robot finger was designed with same structural configuration of the

human hand and the movement of the finger was evaluated via kinematic analysis. Electromyogram (EMG) signals for

hand motions were measured using commercial wearable EMG sensors and classification of hand motions was achieved

by applying the artificial neural network (ANN) algorithm. After training and testing for three kinds of gripping motions via

ANN, it was observed that high classification accuracy can be obtained. A prototype of the proposed robot hand is

manufactured through 3D printing and servomotors are included for position control of fingers. It was demonstrated that

effective realization of gripping motions of the proposed prosthetic robot hand can be achieved by using EMG

measurement and machine learning-based classification under a real-time environment.

Manuscript received: October 15, 2019 / Revised: January 29, 2020 / Accepted: February 10, 2020

This paper was presented at KSPE Spring Conference 2019

1. ์„œ๋ก 

์˜์ˆ˜๋Š” ์ƒ์ง€ ์ ˆ๋‹จ์ž์˜ ์ผ์ƒ์ƒํ™œ์„ ๋•๊ธฐ ์œ„ํ•œ ๋ณด์กฐ ๊ธฐ๊ตฌ๋กœ ํฌ

๊ฒŒ ์žฅ์‹ ์˜์ˆ˜(๋ฏธ์šฉ ์˜์ˆ˜), ์ž‘์—… ์˜์ˆ˜, ์ธ์ฒด ๊ตฌ๋™ํ˜• ์˜์ˆ˜, ์ „๋™ ์˜

์ˆ˜ ๋“ฑ์œผ๋กœ ๋ถ„๋ฅ˜ํ•  ์ˆ˜ ์žˆ๋‹ค. ์žฅ์‹ ์˜์ˆ˜๋Š” ์™ธํ˜•๋งŒ ์†์˜ ๋ชจ์–‘๊ณผ ๋น„

์Šทํ•˜๊ฒŒ ํ•œ ๊ฒƒ์œผ๋กœ ๋™์ž‘์€ ๋ถˆ๊ฐ€๋Šฅํ•œ ๊ฒƒ์ด๊ณ , ์ž‘์—… ์˜์ˆ˜๋Š” ๊ฐˆ๊ณ ๋ฆฌ

๋ชจ์–‘์˜ ์†๋์œผ๋กœ ๋ฌผ๊ฑด์„ ์žก๊ฑฐ๋‚˜ ์ด๋™์‹œํ‚ฌ ์ˆ˜ ์žˆ๋Š” ํ˜•ํƒœ์ด๋‹ค. ์ธ

์ฒด ๊ตฌ๋™ํ˜• ์˜์ˆ˜๋Š” ์–ด๊นจ ์›€์ง์ž„ ๋“ฑ์„ ์ด์šฉํ•ด์„œ ๊ฐ„๋‹จํ•œ ๋™์ž‘์„ ๊ตฌ

ํ˜„ํ•  ์ˆ˜ ์žˆ์œผ๋‚˜, ์ธ์ฒด ํŠน์ • ๋ถ€๋ถ„์— ๋ฌด๋ฆฌ๋ฅผ ์ค„ ์ˆ˜ ์žˆ๋Š” ๋‹จ์ ์ด ์žˆ

๋‹ค. ์ „๋™ ์˜์ˆ˜๋Š” ๋ชจํ„ฐ๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๊ฐ€๋ฝ์„ ์›€์ง์ผ ์ˆ˜ ์žˆ๋Š” ํ˜•

ํƒœ๋กœ ์ œํ•œ๋œ ์†๋™์ž‘์„ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ์ง€๋งŒ, ๊ฐ€๊ฒฉ์ด ๋งค์šฐ ๋†’์•„ ์ œ

ํ•œ์ ์œผ๋กœ ํ™œ์šฉ๋˜๊ณ  ์žˆ๋‹ค. ์ตœ๊ทผ์—๋Š” ์‹ ์ฒด์˜ ์ž”์กด ๊ทผ์œก์—์„œ ์ธก์ •

ํ•  ์ˆ˜ ์žˆ๋Š” ํ‘œ๋ฉด ๊ทผ์ „๋„๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๋™์ž‘์„ ๊ตฌํ˜„ํ•˜๋Š” ๋‹ค์–‘ํ•œ

ํ˜•ํƒœ์˜ ์˜์ˆ˜๋“ค์ด ์ œ์•ˆ๋˜๊ณ  ์žˆ์œผ๋ฉฐ ๊ด€๋ จ ์—ฐ๊ตฌ๊ฐ€ ํ™œ๋ฐœํ•˜๊ฒŒ ์ง„ํ–‰

๋˜๊ณ  ์žˆ๋‹ค.

Saridis์™€ Gootee1๋Š” ์ด๋‘๊ทผ๊ณผ ์‚ผ๋‘๊ทผ์˜ ๊ทผ์ „๋„๋ฅผ ์ธก์ •ํ•˜์—ฌ

์ „์™„์˜ ๋™์ž‘์„ ๊ตฌํ˜„ํ•˜๋Š” ์˜์ˆ˜๋ฅผ ์ œ์•ˆํ•˜๊ณ  ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ๊ฒ€์ฆ

ํ•˜์˜€๋‹ค. Zero Crossing๊ณผ ๋ถ„์‚ฐ์„ ์ด์šฉํ•˜์—ฌ ํŠน์ง•์„ ์ถ”์ถœํ•˜๊ณ , ๊ณ„

์ธต์  ์ง€๋Šฅ ์ œ์–ด๊ธฐ๋ฅผ ์ด์šฉํ•˜์—ฌ ํŒจํ„ด์„ ๋ถ„๋ฅ˜ํ•˜์˜€๋‹ค. Matrone2 ๋“ฑ

์€ ์ฃผ์„ฑ๋ถ„ ๋ถ„์„ ๊ธฐ๋ฐ˜์˜ ์ œ์–ด๊ธฐ๋ฅผ ์ด์šฉํ•œ ๊ณผ์†Œ ๊ตฌ๋™ ์ „๋™ ์˜์ˆ˜์˜

์ œ์–ด๋ฅผ ์ œ์•ˆํ•˜๊ณ , ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํšจ๊ณผ๋ฅผ ๊ณ ์ฐฐํ•˜์˜€๋‹ค. Ma3 ๋“ฑ์€

๊ทผ์œก ์‹œ๋„ˆ์ง€ ๋ชจ๋ธ์„ ์ด์šฉํ•˜์—ฌ ๊ทผ์ „๋„ ๊ธฐ๋ฐ˜ ์˜์ˆ˜์˜ ์†๊ณผ ์†๋ชฉ ์šด

๋™์„ ์ œ์–ดํ•˜๋Š” ๊ฒƒ์„ ์ œ์•ˆํ•˜์˜€๊ณ , ๊ฐ€์ƒ ๊ณต๊ฐ„์—์„œ ์˜์ˆ˜๊ฐ€ ๋™์ž‘์„

ํšจ๊ณผ์ ์œผ๋กœ ๊ตฌํ˜„ํ•˜๋Š” ๊ฒƒ์„ ํ™•์ธํ•˜์˜€๋‹ค. Guo4 ๋“ฑ์€ ๊ทผ์ „๋„ ์‹ ํ˜ธ

์™€ ํ•จ๊ป˜ ๊ทผ์ ์™ธ์„  ๋ถ„๊ด‘๋ฒ•(Near-Infrared Spectroscopy, NIRS)์„

ํ•จ๊ป˜ ์ด์šฉํ•˜์—ฌ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋งŒ์„ ์ด์šฉํ•  ๋•Œ๋ณด๋‹ค ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ

๋†’์ผ ์ˆ˜ ์žˆ๋Š” ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•˜๊ณ  ๊ฐ€์ƒ ๊ณต๊ฐ„์˜ ์˜์ˆ˜๋ฅผ ์ด์šฉํ•˜์—ฌ ๊ฒ€

์ฆํ•˜์˜€๋‹ค. Lee5,6 ๋“ฑ์€ ๊ทผ์ „๋„ ์ธก์ •์„ ์œ„ํ•œ ์ „๊ทน์ด ํฌํ•จ๋˜์–ด ์žˆ

๋Š” ๋‹ˆํŠธ ๋ฐด๋“œ ํ˜•ํƒœ์˜ ์žฅ์น˜๋ฅผ ์ œ์•ˆํ•˜๊ณ , ์‹คํ—˜์œผ๋กœ ํšจ์œจ์„ฑ์„ ๊ฒ€์ฆ

ํ•˜์˜€๋‹ค. Lee7 ๋“ฑ์€ ์˜์ˆ˜์˜ ํŒŒ์ง€ ์„ฑ๋Šฅ ํ‰๊ฐ€ ๋ฐฉ๋ฒ•์„ ์ œ์•ˆํ•˜๊ณ , ์ด๋ฅผ

Copyright ยฉ The Korean Society for Precision Engineering

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/

3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2: Design of Prosthetic Robot Hand and Electromyography-Based

340 / May 2020 ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ

๋ฐ”ํƒ•์œผ๋กœ ์ฃผ์š” ์„ค๊ณ„ ์ธ์ž๋“ค์ด ํŒŒ์ง€ ์„ฑ๋Šฅ์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ์„ ๊ฒ€์ฆ

ํ•˜์˜€๋‹ค. Lim8 ๋“ฑ์€ ์ „๋™ ์˜์ˆ˜์˜ ํŒŒ์ง€๋ ฅ ์ œ์–ด๋ฅผ ์œ„ํ•˜์—ฌ ์ผ€์ด๋ธ”์˜

์žฅ๋ ฅ์„ ์ธก์ •ํ•˜์—ฌ ๋˜๋จน์ž„ ์ œ์–ด๋ฅผ ์ˆ˜ํ–‰ํ•˜๋„๋ก ํ•˜์˜€์œผ๋ฉฐ, ์ง„๋™ ์‹ 

ํ˜ธ๋ฅผ ์ด์šฉํ•œ ํ–…ํ‹ฑ ์‹œ์Šคํ…œ์„ ๊ตฌ์„ฑํ•˜๊ณ  ํ‰๊ฐ€ํ•˜์˜€๋‹ค. Bae9 ๋“ฑ์€ ์ „

๋™ ์˜์ˆ˜ ์‚ฌ์šฉ์ž์—๊ฒŒ ํ–…ํ‹ฑ ํ”ผ๋“œ๋ฐฑ์„ ์ค„ ์ˆ˜ ์žˆ๋„๋ก ๊ฐ๊ฐ์„ ์ธก์ •ํ•˜

๊ณ  ์ „๋‹ฌํ•˜๋Š” ์‹œ์Šคํ…œ์„ ์ œ์•ˆํ•˜๊ณ  ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํšจ์œจ์„ฑ์„ ํ™•์ธ

ํ•˜์˜€๋‹ค. Noh10 ๋“ฑ์€ ๋ฌผ๊ฑด์„ ์žก๋Š” ๋„ค ๊ฐ€์ง€ ๋™์ž‘์— ๋Œ€ํ•˜์—ฌ ์„ธ ๋‹จ

๊ณ„์˜ ์•…๋ ฅ ์กฐ์ ˆ์„ ์ˆ˜ํ–‰ํ•˜๊ธฐ ์œ„ํ•œ ์ถ”์ • ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜๊ณ  ์ •ํ™•๋„

๋ฅผ ์‹คํ—˜์œผ๋กœ ๊ฒ€์ฆํ•˜์˜€๋‹ค. Kim11 ๋“ฑ์€ ๋‡ŒํŒŒ๋ฅผ ์ด์šฉํ•˜์—ฌ ์˜์ˆ˜์˜

์›€์ง์ž„์„ ์ œ์–ดํ•˜๋Š” ๊ฒƒ์„ ์ œ์•ˆํ•˜๊ณ , ๋™์ž‘์— ๋”ฐ๋ฅธ ๋‡ŒํŒŒ์˜ ํŠน์ง•์„

๋ถ„๋ฅ˜ํ•˜๋Š” ๊ธฐ๋ฒ•์„ ์ œ์•ˆํ•˜๊ณ  ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ๊ฒ€์ฆํ•˜์˜€๋‹ค. Jo์™€

Oh12๋Š” ์ตœ๊ทผ ๋‹ค์–‘ํ•œ ๋ถ„์•ผ์— ์ ์šฉ๋˜๊ณ  ์žˆ๋Š” ๋”ฅ๋Ÿฌ๋‹ ๊ธฐ๋ฒ•์„ ์ด์šฉ

ํ•˜์—ฌ ๊ทผ์ „๋„ ์‹ ํ˜ธ์— ๋”ฐ๋ฅธ ์†๋™์ž‘์„ ๋ถ„๋ฅ˜ํ•˜๊ณ  ๋Šฅ๋™ ์˜์ˆ˜์— ์ ์šฉ

์„ ์ œ์•ˆํ•˜์˜€์œผ๋ฉฐ, ๋”ฅ๋Ÿฌ๋‹ ํŒŒ๋ผ๋ฏธํ„ฐ ๋ณ€ํ™”์— ๋Œ€ํ•˜์—ฌ ๋ถ„๋ฅ˜ ์ •ํ™•๋„

๋ฅผ ๊ฒ€์ฆํ•˜์˜€๋‹ค.

๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ์ƒˆ๋กœ์šด ํ˜•ํƒœ์˜ ๋กœ๋ด‡ ์˜์ˆ˜ ๋ชจ๋ธ์„ ์ œ์•ˆํ•˜๊ณ , ์ด

๋ฅผ ์ด์šฉํ•˜์—ฌ ์ผ์ƒ์ƒํ™œ์— ํ•„์š”ํ•œ ๊ธฐ๋ณธ์ ์ธ ์†๋™์ž‘์„ ํšจ๊ณผ์ ์œผ๋กœ

๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ์Œ์„ ํ™•์ธํ•œ๋‹ค. ์‚ฌ๋žŒ์˜ ์†๊ณผ ๋™์ผํ•˜๊ฒŒ ์„ธ ๊ฐœ์˜ ๋ง

ํฌ๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๊ฐ€๋ฝ์„ ๊ตฌ์„ฑํ•˜๊ณ , ์‚ฌ๋žŒ์˜ ํž˜์ค„๊ณผ ๋™์ผํ•œ ์—ญํ• 

์„ ํ•  ์ˆ˜ ์žˆ๋„๋ก ํด๋ฆฌ์—ํ‹ธ๋ Œ ์™€์ด์–ด๋ฅผ ์‚ฝ์ž…ํ•˜๋ฉฐ, ๋ณต์›์„ ์œ„ํ•ด์„œ

ํƒ„์„ฑ์ด ์žˆ๋Š” ๋ฐด๋“œ๋ฅผ ํฌํ•จํ•˜๋„๋ก ํ•œ๋‹ค. ์ œ์•ˆ๋œ ๋งํฌ ๊ตฌ์กฐ๋ฅผ ์ด์šฉ

ํ•˜์—ฌ ์‚ฌ๋žŒ์˜ ์†๊ฐ€๋ฝ ์šด๋™์„ ์œ ์‚ฌํ•˜๊ฒŒ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ๋Š”์ง€ ์†๊ฐ€๋ฝ

๋ ์œ„์น˜์— ๋Œ€ํ•œ ๊ธฐ๊ตฌํ•™ ํ•ด์„์„ ์ˆ˜ํ–‰ํ•œ ํ›„, 3D ํ”„๋ฆฐํ„ฐ๋ฅผ ์ด์šฉํ•˜

์—ฌ ์ œ์ž‘ํ•œ๋‹ค. ์ผ์ƒ์ƒํ™œ์—์„œ ๋งŽ์ด ์“ฐ์ด๋Š” ์‚ฌ๋žŒ์˜ ์†๋™์ž‘ ์„ธ ๊ฐ€์ง€

๋ฅผ ์ •์˜ํ•˜๊ณ , ๋™์ž‘์— ๋Œ€ํ•œ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ธก์ •ํ•œ๋‹ค. ๋™์ผ ๋™์ž‘์—

๋Œ€ํ•˜์—ฌ ๋ฐ˜๋ณต ์ธก์ •๋œ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ด์šฉํ•˜์—ฌ ๋™์ž‘ ์ธ์‹์„ ์œ„ํ•œ

ํ•™์Šต์„ ์ˆ˜ํ–‰ํ•˜๊ณ , ํ…Œ์ŠคํŠธ๋ฅผ ์ˆ˜ํ–‰ํ•˜์—ฌ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋†’์ด๋„๋ก

ํ•œ๋‹ค. ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ด์šฉํ•˜์—ฌ ์‚ฌ์šฉ์ž๊ฐ€ ์˜๋„ํ•˜๋Š” ์†๋™์ž‘์„ ์ธ

์‹ํ•˜๊ณ , ๋กœ๋ด‡ ์˜์ˆ˜๊ฐ€ ๋™์ผํ•œ ๋™์ž‘์„ ์ •ํ™•ํ•˜๊ฒŒ ๊ตฌํ˜„ํ•˜๋Š” ๊ฒƒ์„ ์‹ค

์‹œ๊ฐ„ ํ™˜๊ฒฝ์˜ ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํ™•์ธํ•œ๋‹ค. ์ œ์•ˆ๋œ ๋ฐฉ์‹์ด ์‹ค์ œ ์ƒํ•ญ

์—์„œ๋„ ํšจ๊ณผ์ ์œผ๋กœ ํ™œ์šฉ๋  ์ˆ˜ ์žˆ์Œ์„ ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํ™•์ธํ•œ๋‹ค.

2. ๋กœ๋ด‡ ์˜์ˆ˜ ์„ค๊ณ„

๋ณธ ์—ฐ๊ตฌ์—์„œ ๊ณ ๋ คํ•œ ๋กœ๋ด‡ ์˜์ˆ˜์˜ ์ „์ฒด ๊ตฌ์กฐ ๊ฐœ๋žต๋„๋ฅผ Fig. 1

์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ์‚ฌ๋žŒ์˜ ์†๊ณผ ๋™์ผํ•œ ํฌ๊ธฐ์™€ ๊ตฌ์กฐ์˜ ์†๊ฐ€๋ฝ์„ ๊ฐ–

๋Š” ์˜์ˆ˜๋ฅผ ๊ณ ๋ คํ•˜์˜€์œผ๋ฉฐ, ์—„์ง€์™€ ์—ฐ๊ฒฐ๋œ ์†ํ—ˆ๋ฆฌ๋ผˆ์˜ ์šด๋™ ๊ตฌํ˜„

์„ ์œ„ํ•œ ๋ชจํ„ฐ๋ฅผ ์ œ์™ธํ•œ ๋‚˜๋จธ์ง€ ์†๊ฐ€๋ฝ์˜ ๊ตฌ๋™์„ ์œ„ํ•œ ๋ชจํ„ฐ๋“ค์€

์†๋ชฉ ์•„๋ž˜์ชฝ์— ์œ„์น˜ํ•˜๋„๋ก ํ•˜์˜€๋‹ค. ์‹ค์ œ ์‚ฌ๋žŒ์—๊ฒŒ ์žฅ์ฐฉ์„ ์œ„ํ•œ

์†Œ์ผ“๊ณผ ๊ฐ™์€ ์—ฐ๊ฒฐ ๋ถ€๋ถ„์€ ๊ณ ๋ คํ•˜์ง€ ์•Š๊ณ  ์†๊ฐ€๋ฝ์˜ ๋™์ž‘ ๊ตฌํ˜„๋งŒ

์„ ๊ณ ๋ คํ•˜์—ฌ ์„ค๊ณ„๋ฅผ ์ˆ˜ํ–‰ํ•˜์˜€๋‹ค. ๋กœ๋ด‡ ์˜์ˆ˜์˜ ํฌ๊ธฐ๋Š” ์„ฑ์ธ ๋‚จ์„ฑ

์˜ ์™ผ์†์„ ๊ธฐ์ค€์œผ๋กœ ํ•˜์˜€์œผ๋ฉฐ, ์†๋ฐ”๋‹ฅ์˜ ๋†’์ด์™€ ํญ์€ 81 mm๋กœ

์„ค๊ณ„ํ•˜์˜€๋‹ค. ์‚ฌ๋žŒ์˜ ์†์—๋Š” ์—„์ง€, ๊ฒ€์ง€, ์ค‘์ง€, ์•ฝ์ง€, ์†Œ์ง€์˜ ๋‹ค์„ฏ

๊ฐœ์˜ ์†๊ฐ€๋ฝ์ด ์žˆ์œผ๋ฉฐ, ์—„์ง€๋ฅผ ์ œ์™ธํ•œ ๋‚˜๋จธ์ง€ ๋„ค ์†๊ฐ€๋ฝ์€ ๋๋งˆ

๋””๋ผˆ, ์ค‘๊ฐ„๋งˆ๋””๋ผˆ, ์ฒซ๋งˆ๋””๋ผˆ, ์„ธ ๊ฐœ์˜ ์†๊ฐ€๋ฝ๋ผˆ๋กœ ๊ตฌ์„ฑ๋˜์–ด ์žˆ๊ณ ,

์—„์ง€๋Š” ์ค‘๊ฐ„๋งˆ๋””๋ผˆ๊ฐ€ ์—†์ด ๋๋งˆ๋””๋ผˆ์™€ ์ฒซ๋งˆ๋””๋ผˆ๋กœ๋งŒ ๊ตฌ์„ฑ๋˜์–ด

์žˆ๋‹ค. ๊ฐ ์†๊ฐ€๋ฝ์˜ ์ฒซ๋งˆ๋””๋ผˆ๋Š” ์†๋ฐ”๋‹ฅ ๋‚ด๋ถ€์— ์žˆ๋Š” ์†ํ—ˆ๋ฆฌ๋ผˆ๋ฅผ

ํ†ตํ•ด ์†๋ชฉ๋ผˆ์™€ ์—ฐ๊ฒฐ๋˜์–ด ์žˆ๋‹ค. ์ด๋Ÿฌํ•œ ์†๊ฐ€๋ฝ์˜ ํ•ด๋ถ€ํ•™์  ํŠน์ง•

์„ ๊ณ ๋ คํ•˜์—ฌ ์„ธ ๊ฐœ์˜ ๋งํฌ ๊ธฐ๊ตฌ๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๊ฐ€๋ฝ์„ ๊ตฌ์„ฑํ•˜์˜€

๋‹ค. ๋ณธ ์—ฐ๊ตฌ์—์„œ ์ œ์•ˆ๋œ ์†๊ฐ€๋ฝ์˜ ๋ชจ๋ธ์„ Fig. 2(a)์— ๋‚˜ํƒ€๋‚ด์—ˆ

๊ณ , ์†๊ฐ€๋ฝ ๋ชจ๋ธ์˜ ๋‹จ๋ฉด์„ Fig. 2(b)์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ์—„์ง€๋ฅผ ์†๋ฐ”

๋‹ฅ ์•ˆ์ชฝ์œผ๋กœ ๊ตฝํžˆ๋Š” ๋™์ž‘์€ ์†๋ฐ”๋‹ฅ ๋‚ด๋ถ€์— ์žˆ๋Š” ์†ํ—ˆ๋ฆฌ๋ผˆ์— ์˜

ํ•˜์—ฌ ๊ตฌํ˜„๋˜๋ฏ€๋กœ, ์ œ์•ˆ๋œ ์˜์ˆ˜์—์„œ ์—„์ง€๋Š” ์†ํ—ˆ๋ฆฌ๋ผˆ๋ฅผ ํฌํ•จํ•˜์—ฌ

์„ธ ๊ฐœ์˜ ๋งํฌ๋ฅผ ์ด์šฉํ•˜์—ฌ ๊ตฌ์„ฑํ•˜์˜€๋‹ค. ์ œ์•ˆ๋œ ์†๊ฐ€๋ฝ ๋ชจ๋ธ์—๋Š”

๊ฐ ๊ด€์ ˆ ๊ตฌ๋™์„ ์œ„ํ•œ ๋ชจํ„ฐ๋ฅผ ๋”ฐ๋กœ ์‚ฌ์šฉํ•˜์ง€ ์•Š๊ณ , ์™€์ด์–ด๋ฅผ ์ด

์šฉํ•˜์—ฌ ํ•œ ๋ฒˆ์˜ ๋ชจํ„ฐ ๊ตฌ๋™์œผ๋กœ ์†๊ฐ€๋ฝ์˜ ์„ธ ๊ฐœ์˜ ๋งํฌ๊ฐ€ ์›€์ง์ผ

์ˆ˜ ์žˆ๋„๋ก ํ•˜์˜€๋‹ค. Fig. 2(b)์˜ ๋‹จ๋ฉด๋„์— ๋‚˜ํƒ€๋‚œ ๋ฐ”์™€ ๊ฐ™์ด ์†๊ฐ€

๋ฝ ๋‚ด๋ถ€์— ํด๋ฆฌ์—ํ‹ธ๋ Œ ์™€์ด์–ด๋ฅผ ์„ค์น˜ํ•˜์—ฌ ์†๊ฐ€๋ฝ์„ ๊ตฌ์„ฑํ•˜๋Š”

์„ธ ๊ฐœ์˜ ๋งํฌ๋ฅผ ์—ฐ๊ฒฐํ•˜๊ณ , ์†๊ฐ€๋ฝ์˜ ํž˜์ค„๊ณผ ๊ฐ™์€ ์—ญํ• ์„ ์ˆ˜ํ–‰ํ•˜

๋„๋ก ํ•˜์˜€์œผ๋ฉฐ, ์†๊ฐ€๋ฝ์˜ ๊ตฝํž˜ ๋™์ž‘ ํ›„ ์ดˆ๊ธฐ ์œ„์น˜๋กœ ๋ณต๊ท€๋ฅผ ์œ„

ํ•ด์„œ ํƒ„์„ฑ ๋ฐด๋“œ๋ฅผ ์ด์šฉํ•˜์˜€๋‹ค. ๋‹จ๋ฉด๋„์—์„œ ์†๊ฐ€๋ฝ์˜ ์•ˆ์ชฝ์— ํž˜

์ค„์ด, ๋ฐ”๊นฅ์ชฝ์— ํƒ„์„ฑ ๋ฐด๋“œ๊ฐ€ ์žฅ์ฐฉ๋œ๋‹ค. ์ œ์•ˆ๋œ ์†๊ฐ€๋ฝ ๋ชจ๋ธ์ด

Fig. 1 Structural configuration of the prosthetic robot hand

Fig. 2 The proposed finger model

Page 3: Design of Prosthetic Robot Hand and Electromyography-Based

ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ May 2020 / 341

์‚ฌ๋žŒ์˜ ์†๊ณผ ๋™์ผํ•œ ์šด๋™์„ ๊ตฌํ˜„ํ•˜๋Š”์ง€ ํ™•์ธํ•˜๊ธฐ ์œ„ํ•˜์—ฌ ์†๊ฐ€

๋ฝ ๋์ ์˜ ์šด๋™ ๊ฒฝ๋กœ๋ฅผ ํ™•์ธํ•˜์˜€๊ณ , ์—„์ง€ ์†๊ฐ€๋ฝ๊ณผ ๋‚˜๋จธ์ง€ ๋„ค

์†๊ฐ€๋ฝ์˜ ์‹œ๋ฎฌ๋ ˆ์ด์…˜ ๊ฒฐ๊ณผ๋ฅผ ์‹ค์ œ ์‚ฌ๋žŒ ์†์˜ ์‚ฌ์ง„๊ณผ ํ•จ๊ป˜ Figs.

3(a)์™€ 3(b)์— ๊ฐ๊ฐ ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ๋‹ค์„ฏ ์†๊ฐ€๋ฝ ๋ชจ๋‘ ์‚ฌ๋žŒ ์†์˜ ์šด

๋™๊ณผ ๋™์ผํ•œ ํ˜•ํƒœ๋กœ ์šด๋™ํ•˜๋Š” ๊ฒƒ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ์œผ๋ฉฐ, ์ œ์•ˆ๋œ

์†๊ฐ€๋ฝ ๋ชจ๋ธ์„ ์˜์ˆ˜์— ์ ์šฉํ•˜๋Š” ๊ฒƒ์ด ๊ฐ€๋Šฅํ•จ์„ ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค.

3. ์†๋™์ž‘ ์ธ์‹

๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ด์šฉํ•˜์—ฌ ์†๋™์ž‘์„ ์ธ์‹ํ•˜๊ณ 

๊ทธ ๊ฒฐ๊ณผ๋ฅผ ์˜์ˆ˜๋ฅผ ํ†ตํ•ด ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•˜์˜€๋‹ค. ๊ทผ์ „๋„ ์ธก์ •

์€ ์•”๋ฐด๋“œ ํ˜•ํƒœ์˜ ์ƒ์šฉ ๊ทผ์ „๋„ ์ธก์ • ์žฅ์น˜์ธ Talmic์‚ฌ์˜ MYO๋ฅผ

์ด์šฉํ•˜์˜€๋‹ค. MYO์—๋Š” 8 ์ฑ„๋„์˜ ๊ทผ์ „๋„ ์„ผ์„œ๊ฐ€ ์žฅ์ฐฉ๋˜์–ด ์žˆ์œผ

๋ฉฐ, ์ƒ˜ํ”Œ๋ง ์ฃผํŒŒ์ˆ˜๋Š” 200 Hz๋กœ ๊ณ ์ •๋˜์–ด ์žˆ๋‹ค. ์‹ค์ œ ๊ทผ์ „๋„ ์‹ 

ํ˜ธ๋Š” mV ๋‹จ์œ„๋กœ ์ธก์ •๋˜๋‚˜ MYO๋Š” -128์—์„œ 128 ์‚ฌ์ด์˜ ์ •์ˆ˜

๊ฐ’์œผ๋กœ ๊ฒฐ๊ณผ๋ฅผ ์ถœ๋ ฅํ•˜๋ฉฐ ๋‹จ์œ„๋Š” Activation์ด๋‹ค. ์ธก์ •๋œ ๊ทผ์ „๋„

์‹ ํ˜ธ๋Š” ๋ธ”๋ฃจํˆฌ์Šค ํ†ต์‹ ์„ ํ†ตํ•˜์—ฌ ๊ฐœ์ธ์šฉ ์ปดํ“จํ„ฐ๋กœ ์ „์†ก๋˜๊ณ ,

MATLAB์„ ์ด์šฉํ•˜์—ฌ ์‹ค์‹œ๊ฐ„์œผ๋กœ ์ธก์ •๋œ ๊ทผ์ „๋„ ๋ฐ์ดํ„ฐ๋ฅผ ์ˆ˜

์ง‘ํ•˜๊ณ , ๋ถ„๋ฅ˜๋ฅผ ์ˆ˜ํ–‰ํ•˜๋„๋ก ํ•˜์˜€๋‹ค. ๋ณธ ์—ฐ๊ตฌ์—์„œ ์ˆ˜ํ–‰๋œ ๋ฐ์ดํ„ฐ

์ˆ˜์ง‘๊ณผ ๋ถ„๋ฅ˜์˜ ๋™์ž‘ ์ธ์‹ ๊ณผ์ •์„ Fig. 4์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค.

์†๊ฐ€๋ฝ์„ ๊ตฝํž ๋•Œ ์‚ฌ์šฉํ•˜๋Š” ์‹ฌ์ง€๊ตด๊ทผ(Flexor Digitorum

Profundus), ์ฒœ์ง€๊ตด๊ทผ(Flexor Digitorum Superficialis), ์žฅ๋ฌด์ง€๊ตด๊ทผ

(Flexor Pollicis Longus), ์†๊ฐ€๋ฝ์„ ํŽผ ๋•Œ ์‚ฌ์šฉํ•˜๋Š” ์ˆ˜์ง€์‹ ๊ทผ

(Extensor Digitorum), ์žฅ๋ฌด์ง€์‹ ๊ทผ(Extensor Pollicis Longus), ์žฅ

๋ฌด์ง€์™ธ์ „๊ทผ(Abductor Pollicis Longus) ๋“ฑ์˜ ๊ทผ์œก์ด ์œ„์น˜ํ•˜๊ณ  ์žˆ

๋Š” ์•„๋ž˜ํŒ”(์ „์™„)์˜ ํŒ”๊ฟˆ์น˜ ์•„๋ž˜์ชฝ์— ๊ทผ์ „๋„ ์ธก์ • ์žฅ์น˜๋ฅผ ์ฐฉ์šฉํ•˜

๊ณ  ๋ฐ์ดํ„ฐ๋ฅผ ์ˆ˜์ง‘ํ•˜์˜€๋‹ค. ๊ธฐ๊ณ„ ํ•™์Šต์„ ์œ„ํ•ด์„œ๋Š” ์†๋™์ž‘์— ๋Œ€ํ•˜

์—ฌ ์ธก์ •๋œ ๊ทผ์ „๋„ ์‹ ํ˜ธ์˜ ํŠน์ง•์„ ์ถ”์ถœํ•ด์•ผ ํ•˜๋Š”๋ฐ, ๋ณธ ์—ฐ๊ตฌ์—์„œ

๋Š” ์ ˆ๋Œ€ ํ‰๊ท ๊ฐ’(Mean Absolute Value, MAV)์„ ํŠน์ง•์œผ๋กœ ์‚ฌ์šฉํ•˜

์˜€๋‹ค. ์ ˆ๋Œ€ ํ‰๊ท ๊ฐ’์€ ๋‹ค์Œ ์‹(1)๊ณผ ๊ฐ™์€ ์‹์œผ๋กœ ๋‚˜ํƒ€๋‚ผ ์ˆ˜ ์žˆ๋‹ค.

(1)

์—ฌ๊ธฐ์„œ, x(i)๋Š” ์ธก์ •๋œ ๋ฐ์ดํ„ฐ์ด๊ณ , L์€ ์ธก์ • ์ฃผํŒŒ์ˆ˜์™€ ์œˆ๋„์ž‰

(Windowing) ์‹œ๊ฐ„์˜ ๊ณฑ์œผ๋กœ ๊ฒฐ์ •๋˜๋Š” ์œˆ๋„์šฐ์˜ ํฌ๊ธฐ๋กœ ๋ณธ ์—ฐ๊ตฌ

์—์„œ๋Š” L = 100์œผ๋กœ ์‚ฌ์šฉํ•˜์˜€๋‹ค.

๋‹ค์–‘ํ•œ 8ํšŒ์˜ ์†๋™์ž‘์— ๋Œ€ํ•˜์—ฌ ์ธก์ •๋œ ๊ทผ์ „๋„์˜ ์›์‹ ํ˜ธ์™€ ์ ˆ

๋Œ€ํ‰๊ท ๊ฐ’์„ Figs. 5(a)์™€ 5(b)์— ๊ฐ๊ฐ ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ํ•œ ๊ฐ€์ง€ ์†๋™

์ž‘์— ๋Œ€ํ•˜์—ฌ ๋™์‹œ์— 8 ์ฑ„๋„์˜ ๊ทผ์ „๋„ ์‹ ํ˜ธ๊ฐ€ ์ธก์ •๋˜๋ฏ€๋กœ Fig.

5(c)์™€ ๊ฐ™์ด ๋‚˜ํƒ€๋‚˜๊ณ , ๋ฐ์ดํ„ฐ๋Š” 100X8 ํ˜•ํƒœ์˜ ํ–‰๋ ฌ๋กœ ์ €์žฅ๋œ๋‹ค.

์ด๋Ÿฌํ•œ ๋ฐ์ดํ„ฐ๋ฅผ ์ฒ˜๋ฆฌํ•˜๊ธฐ ์‰ฝ๋„๋ก Fig. 5(d)์™€ ๊ฐ™์ด 1์—์„œ 8๊นŒ

์ง€ ์ฑ„๋„์˜ ๊ฒฐ๊ณผ๋ฅผ ์ด์–ด ๋ถ™์—ฌ์„œ 800X1 ํ˜•ํƒœ์˜ ์—ด๋ฒกํ„ฐ๋กœ ๋ณ€ํ™˜ํ•˜

๊ณ , ์ด๋ฅผ ํ•ด๋‹น ๋™์ž‘์— ๋Œ€ํ•œ ํŠน์ง• ๋ฒกํ„ฐ๋กœ ์‚ฌ์šฉํ•˜๊ฒŒ ๋œ๋‹ค. ๋ณธ ์—ฐ๊ตฌ

์—์„œ๋Š” ์ผ์ƒ์ƒํ™œ์—์„œ ๊ฐ€์žฅ ๋งŽ์ด ์‚ฌ์šฉํ•˜๋Š” ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ(Cylindrical

Grasp), ํ•€์น˜ ๊ทธ๋ฆฝ(Pinch Grasp), ๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ(Lateral Grasp)์˜

์„ธ ๊ฐ€์ง€ ๋™์ž‘์— ๋Œ€ํ•˜์—ฌ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ๋„๋ก ํ•˜์˜€๋‹ค. ๋ฌผ๋ณ‘ ๋“ฑ์˜ ๋ฌผ

๊ฑด์„ ์žก์„ ๋•Œ ์‚ฌ์šฉ๋˜๋Š” ๋™์ž‘์ด ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ, ์—„์ง€์™€ ๊ฒ€์ง€๋ฅผ ์ด

์šฉํ•˜์—ฌ ์ž‘์€ ๋ฌผ์ฒด๋ฅผ ์žก๋Š” ๋™์ž‘์ด ํ•€์น˜ ๊ทธ๋ฆฝ, ๋ช…ํ•จ๊ณผ ๊ฐ™์ด ์–‡์€

์ข…์ด ๋“ฑ์„ ์žก๋Š” ๋™์ž‘์ด ๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ์ด๋‹ค. ์„ธ ๊ฐ€์ง€ ๋™์ž‘๊ณผ ํ•ด๋‹น

ํŠน์ง• ๋ฒกํ„ฐ๋ฅผ Fig. 6์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. Fig. 6(a)์˜ ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ์€ ๋ฌผ

๊ฑด์„ ์žก๋Š” ํ˜•ํƒœ๋กœ ๊ตฌํ˜„๋˜์ง€๋งŒ, ๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ์ฃผ๋จน์„ ์ฅ๋Š” ํ˜•ํƒœ

์˜ ์†๋™์ž‘์„ ์ธ์‹ํ•˜์—ฌ ๋กœ๋ด‡ ์˜์ˆ˜๊ฐ€ ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ์„ ๊ตฌํ˜„ํ•˜๋„๋ก

ํ•˜์˜€๋‹ค.

ํ•œ ๊ฐ€์ง€ ๋™์ž‘์„ 160ํšŒ ๋ฐ˜๋ณต ์ธก์ •ํ•˜์—ฌ ์ด 480๊ฐœ์˜ ๋ฐ์ดํ„ฐ๋ฅผ

ํ™•๋ณดํ•˜์˜€๊ณ , ์ด ์ค‘์—์„œ 300๊ฐœ๋Š” ํ•™์Šต์— ์‚ฌ์šฉํ•˜๊ณ , ๋‚˜๋จธ์ง€ 180๊ฐœ

๋Š” ํ…Œ์ŠคํŠธ์— ์‚ฌ์šฉํ•˜๋„๋ก ํ•˜์˜€๋‹ค. ์†๋™์ž‘ ์ธ์‹์„ ์œ„ํ•œ ๊ธฐ๊ณ„ ํ•™

์Šต์€ ์˜ค๋ฅ˜ ์—ญ์ „ํŒŒ ์•Œ๊ณ ๋ฆฌ์ฆ˜์„ ์ด์šฉํ•˜์—ฌ ํ•™์Šต์„ ์ง„ํ–‰ํ•˜๋Š” ์ธ๊ณต

MAV1

L--- x i L

i=1=Fig. 3 Simulation results for locus of fingertips

Fig. 4 Process of gesture recognition

Page 4: Design of Prosthetic Robot Hand and Electromyography-Based

342 / May 2020 ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ

์‹ ๊ฒฝ๋ง(Artificial Neural Network)์˜ ๋‹ค์ธต ํผ์…‰ํŠธ๋ก (Multi-Layer

Perceptron) ๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์˜€๊ณ , MATLAB์˜ Neural Network

Toolbox๋ฅผ ์ด์šฉํ•˜์—ฌ ๊ตฌํ˜„ํ•˜์˜€๋‹ค. ๋‹ค์ธต ํผ์…‰ํŠธ๋ก ์€ ๊ฐ๊ฐ ์„ธ ๊ฐœ์˜

๋‰ด๋Ÿฐ์œผ๋กœ ๊ตฌ์„ฑ๋˜๋Š” ์ž…๋ ฅ์ธต๊ณผ ์ถœ๋ ฅ์ธต์„ ๊ฐ–๊ณ  ์žˆ์œผ๋ฉฐ, ์€๋‹‰์ธต์˜

๋‰ด๋Ÿฐ์˜ ์ˆ˜๋ฅผ ๋ณ€ํ™”์‹œ์ผœ๊ฐ€๋ฉด์„œ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ํ™•์ธํ•˜์˜€๊ณ , ์ด๋ฅผ

Fig. 7์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ์ดˆ๊ธฐ์—๋Š” ์€๋‹‰ ๋‰ด๋Ÿฐ์˜ ์ˆ˜๊ฐ€ ์ฆ๊ฐ€ํ• ์ˆ˜๋ก ๋ถ„

๋ฅ˜ ์ •ํ™•๋„๊ฐ€ ์ฆ๊ฐ€ํ•˜์ง€๋งŒ, 10๊ฐœ ์ดํ›„์—๋Š” ์ˆ˜๋ ดํ•˜๋Š” ๊ฒƒ์„ ํ™•์ธํ• 

์ˆ˜ ์žˆ๋‹ค. Fig. 8(a)์—๋Š” 300๊ฐœ์˜ ๋ฐ์ดํ„ฐ๋ฅผ ์ด์šฉํ•œ ํ•™์Šต ๊ณผ์ •์—์„œ

์˜ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋‚˜ํƒ€๋‚ด์—ˆ๊ณ , Fig. 8(b)์—๋Š” ํ…Œ์ŠคํŠธ ๊ณผ์ •์˜ ๋ถ„

๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. Fig. 8์—์„œ ์ดˆ๋ก์ƒ‰์œผ๋กœ ํ‘œ์‹œ๋œ ๋ถ€๋ถ„์€

๋ชฉํ‘œ ๋™์ž‘์— ๋Œ€ํ•ด์„œ ๋ถ„๋ฅ˜๊ฐ€ ์ •ํ™•ํ•˜๊ฒŒ ๋œ ๊ฒฝ์šฐ์ด๊ณ , ๋ถ‰์€์ƒ‰์œผ๋กœ

ํ‘œ์‹œ๋œ ๋ถ€๋ถ„์€ ๋ชฉํ‘œ ๋™์ž‘๊ณผ ๋ถ„๋ฅ˜๋œ ๋™์ž‘์ด ์ผ์น˜ํ•˜์ง€ ์•Š๋Š” ๋ถ„๋ฅ˜

๊ฐ€ ์ž˜๋ชป๋œ ๊ฒฝ์šฐ๋ฅผ ๋‚˜ํƒ€๋‚ธ๋‹ค. ํ•™์Šต ๊ณผ์ •์—์„œ ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ๊ณผ ํ•€์น˜

๊ทธ๋ฆฝ์€ ๊ฐ๊ฐ 99%์™€ 100%์˜ ๋งค์šฐ ๋†’์€ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋ณด์˜€์œผ๋‚˜,

๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ์€ ์ƒ๋Œ€์ ์œผ๋กœ ๋‚ฎ์€ 85%์˜ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๊ฐ–๋Š” ๊ฒƒ์„Fig. 5 Measured EMG data

Fig. 6 Hand motions and corresponding EMG feature vectors

Page 5: Design of Prosthetic Robot Hand and Electromyography-Based

ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ May 2020 / 343

ํ™•์ธํ•  ์ˆ˜ ์žˆ๋‹ค. ๊ฐ ๋™์ž‘๋ณ„ 60๊ฐœ, ์ด 180๊ฐœ์˜ ๋ฐ์ดํ„ฐ๋ฅผ ์ด์šฉํ•œ

ํ…Œ์ŠคํŠธ ๊ณผ์ •์—์„œ๋„ ํ•€์น˜ ๊ทธ๋ฆฝ์€ 100%์˜ ๋งค์šฐ ๋†’์€ ๋ถ„๋ฅ˜ ์ •ํ™•๋„

๋ฅผ ๋ณด์˜€์œผ๋‚˜, ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ์€ ๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ์œผ๋กœ ์ž˜๋ชป ๋ถ„๋ฅ˜๋˜๋Š” ๊ฒฝ

์šฐ๊ฐ€ 3ํšŒ ๋ฐœ์ƒํ•˜์—ฌ 95%์˜ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. ๋ž˜ํ„ฐ๋Ÿด

๊ทธ๋ฆฝ์€ ์ƒ๋Œ€์ ์œผ๋กœ ๋” ๋‚ฎ์€ 86.7%์˜ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋‚˜ํƒ€๋‚ด์—ˆ๊ณ ,

์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ ๋™์ž‘์œผ๋กœ ์ž˜๋ชป ๋ถ„๋ฅ˜๋˜๋Š” ๊ฒฝ์šฐ๊ฐ€ ๋งŽ์€ ๊ฒƒ์œผ๋กœ ๋‚˜

ํƒ€๋‚ฌ๋‹ค. ์ด๋Š” ์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ๊ณผ ๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ์˜ ๊ฒฝ์šฐ, ์—„์ง€์™€ ๊ฒ€์ง€

์†๊ฐ€๋ฝ์˜ ๋™์ž‘์€ ๋‹ค์†Œ ์ฐจ์ด๊ฐ€ ์žˆ์œผ๋‚˜, ๋‚˜๋จธ์ง€ ์„ธ ์†๊ฐ€๋ฝ์€ ๋™

์ž‘์ด ๋งค์šฐ ์œ ์‚ฌํ•˜๊ธฐ ๋•Œ๋ฌธ์ธ ๊ฒƒ์œผ๋กœ ํŒ๋‹จ๋œ๋‹ค. ์ด๋Š” Fig. 8(a)์˜

ํ•™์Šต ๊ฒฐ๊ณผ์™€ ์œ ์‚ฌํ•œ ๊ฒฝํ–ฅ์„ ๊ฐ–๋Š” ๊ฒƒ์œผ๋กœ ํŒ๋‹จ๋œ๋‹ค.

4. ์†๋™์ž‘ ๊ตฌํ˜„

3D ํ”„๋ฆฐํ„ฐ๋ฅผ ์ด์šฉํ•˜์—ฌ ์ œ์•ˆ๋œ ์˜์ˆ˜ ํ˜•์ƒ์„ ์ถœ๋ ฅํ•˜๊ณ , ์†๊ฐ€๋ฝ

๋™์ž‘ ๊ตฌํ˜„์„ ์œ„ํ•œ ๋ชจํ„ฐ๋ฅผ ์žฅ์ฐฉํ•˜์—ฌ ๋กœ๋ด‡ ์˜์ˆ˜๋ฅผ ์ œ์ž‘ํ•˜์˜€์œผ๋ฉฐ,

Fig. 9์— ๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. FDM (Fused Deposition Modeling) ๋ฐฉ์‹์˜

3D ํ”„๋ฆฐํ„ฐ๋ฅผ ์‚ฌ์šฉํ•˜์˜€์œผ๋ฉฐ, ์นœํ™˜๊ฒฝ ์ˆ˜์ง€์ธ PLA (Poly Lactic

Acid) ํ•„๋ผ๋ฉ˜ํŠธ๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์—„์ง€ ์†๊ฐ€๋ฝ์˜ ์ฒซ๋งˆ๋””๋ผˆ ๋™์ž‘์„

์œ„ํ•œ ๋ชจํ„ฐ๋Š” ์†๋ฐ”๋‹ฅ ๋‚ด๋ถ€์— ์„ค์น˜ํ•˜์˜€๊ณ , ์—„์ง€, ๊ฒ€์ง€, ์ค‘์ง€, ์•ฝ์ง€

๋ฐ ์†Œ์ง€ ๋™์ž‘์„ ์œ„ํ•˜์—ฌ ๋„ค ๊ฐœ์˜ ๋ชจํ„ฐ๋ฅผ ์†๋ชฉ ์•„๋žซ๋ถ€๋ถ„์— ์„ค์น˜ํ•˜

์˜€๋‹ค. ํž˜์ค„์€ ํด๋ฆฌ์—ํ‹ธ๋ Œ ์žฌ์งˆ์˜ ์ง€๋ฆ„ 0.37 mm ์™€์ด์–ด๋ฅผ ์‚ฌ์šฉ

ํ•˜์˜€๊ณ , ํƒ„์„ฑ ๋ฐด๋“œ๋Š” ํด๋ฆฌ์—์Šคํ…Œ๋ฅด ์žฌ์งˆ์˜ ์ง€๋ฆ„ 3.5 mm ์™€์ด์–ด

๋ฅผ ์‚ฌ์šฉํ•˜์˜€๋‹ค. ์ œ์•ˆ๋œ ๋กœ๋ด‡ ์˜์ˆ˜์˜ ์‹ค์‹œ๊ฐ„ ๋™์ž‘ ๊ตฌํ˜„ ์„ฑ๋Šฅ์„

ํ‰๊ฐ€ํ•˜๊ธฐ ์œ„ํ•˜์—ฌ Fig. 10๊ณผ ๊ฐ™์ด ์‹คํ—˜์„ ๊ตฌ์„ฑํ•˜์˜€๋‹ค. MYO๋ฅผ

์ด์šฉํ•˜์—ฌ ์‚ฌ๋žŒ์˜ ์†๋™์ž‘์— ๋Œ€ํ•œ ๊ทผ์ „๋„๋ฅผ ์ธก์ •ํ•˜๊ณ  ๋ธ”๋ฃจํˆฌ์Šค

ํ†ต์‹ ์„ ์ด์šฉํ•˜์—ฌ ์ปดํ“จํ„ฐ๋กœ ์ „๋‹ฌํ•œ๋‹ค. ์ „๋‹ฌ๋œ ์‹ ํ˜ธ๋ฅผ ๋ฐ”ํƒ•์œผ๋กœ

MATLAB์„ ์ด์šฉํ•˜์—ฌ ์‹ค์‹œ๊ฐ„์œผ๋กœ ๋™์ž‘์„ ๋ถ„๋ฅ˜ํ•œ๋‹ค. ๋ถ„๋ฅ˜๋œ ๋™์ž‘

์— ๋Œ€ํ•˜์—ฌ ์†๊ฐ€๋ฝ ๋์˜ ์œ„์น˜๋ฅผ ๊ฒฐ์ •ํ•˜๊ณ  ์ด๋ฅผ ์œ ์„  USB ์žฅ์น˜๋ฅผ

ํ†ตํ•˜์—ฌ ์•„๋‘์ด๋…ธ UNO ๋ณด๋“œ๋กœ ์ „์†กํ•˜๋ฉฐ, ์•„๋‘์ด๋…ธ ๋ณด๋“œ์—์„œ๋Š”

Fig. 7 Classification accuracy according to number of hidden

neurons

Fig. 8 Classification accuracy for three hand motions with ten

hidden neurons

Fig. 9 Manufactured prosthetic hand

Fig. 10 Experimental setup for real time hand motion realization

Page 6: Design of Prosthetic Robot Hand and Electromyography-Based

344 / May 2020 ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ

๋ชจํ„ฐ์˜ ์ž…๋ ฅ ์‹ ํ˜ธ๋ฅผ ์ธ๊ฐ€ํ•˜๊ฒŒ ๋œ๋‹ค.

๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ด์šฉํ•˜์—ฌ ์‚ฌ๋žŒ์˜ ์†๋™์ž‘์„ ์ธ์‹ํ•˜๊ณ , ์ด๋ฅผ ์‹ค

์‹œ๊ฐ„ ํ™˜๊ฒฝ์—์„œ ์˜์ˆ˜๋ฅผ ๋™์ž‘ํ•˜๋„๋ก ํ•œ ์‹คํ—˜์˜ ์‚ฌ์ง„์„ Fig. 11์—

๋‚˜ํƒ€๋‚ด์—ˆ๋‹ค. Fig. 11(a)๋Š” ์ดˆ๊ธฐ ๊ธฐ๋ณธ ๋™์ž‘ ์ƒํƒœ์ด๋‹ค. Fig. 11(b)๋Š”

์‹ค๋ฆฐ๋” ๊ทธ๋ฆฝ์„ ์ด์šฉํ•˜์—ฌ ๋ฌผ์ฒด๋ฅผ ์žก๋Š” ๋™์ž‘, Fig. 11(c)๋Š” ํ•€์น˜

๊ทธ๋ฆฝ์„ ์ด์šฉํ•˜์—ฌ ํƒ๊ตฌ๊ณต๊ณผ ๊ฐ™์€ ์ž‘์€ ๋ฌผ์ฒด๋ฅผ ์žก๋Š” ๋™์ž‘, Fig.

11(d)๋Š” ๋ž˜ํ„ฐ๋Ÿด ๊ทธ๋ฆฝ์„ ์ด์šฉํ•˜์—ฌ ์–‡์€ ๋ฌผ์ฒด๋ฅผ ์žก๋Š” ๋™์ž‘์„ ๊ฐ๊ฐ

๊ตฌํ˜„ํ•œ ๊ฒƒ์ด๋‹ค. ์‹ค์‹œ๊ฐ„ ํ™˜๊ฒฝ์—์„œ ์‚ฌ๋žŒ ์†์˜ ๋™์ž‘์„ ์ธ์‹ํ•˜๊ณ  ๋กœ

๋ด‡ ์˜์ˆ˜์˜ ๋™์ž‘์„ ๋ช…ํ™•ํ•˜๊ฒŒ ๊ตฌํ˜„ํ•˜์˜€์œผ๋ฉฐ ์‹ค์ œ๋กœ ๋ฌผ๊ฑด์„ ์žก๊ฑฐ

๋‚˜ ๋“ค ์ˆ˜ ์žˆ๋Š” ๊ฒƒ์„ ํ™•์ธํ•˜์˜€๋‹ค.

5. ๊ฒฐ๋ก 

๋ณธ ์—ฐ๊ตฌ์—์„œ๋Š” ๋กœ๋ด‡ ์˜์ˆ˜์— ๋Œ€ํ•œ ์„ค๊ณ„, ํ•ด์„, ์ œ์ž‘์„ ์ˆ˜ํ–‰ํ•˜

๊ณ  ์ธก์ •๋œ ๊ทผ์ „๋„ ์‹ ํ˜ธ์— ๋Œ€ํ•œ ์‹ค์‹œ๊ฐ„ ์†๋™์ž‘ ๊ตฌํ˜„ ์„ฑ๋Šฅ์„ ์‹คํ—˜

์„ ํ†ตํ•˜์—ฌ ์ œ์•ˆ๋œ ์‹œ์Šคํ…œ์˜ ์„ฑ๋Šฅ์„ ํ‰๊ฐ€ํ•˜์˜€๋‹ค. ์‚ฌ๋žŒ์˜ ์†๊ฐ€๋ฝ

๊ณผ ๋™์ผํ•œ ๊ตฌ์กฐ์  ํŠน์ง•์„ ๊ฐ–๋Š” ์˜์ˆ˜๋ฅผ ์„ค๊ณ„ํ•˜๊ณ , ๊ธฐ๊ตฌํ•™ ํ•ด์„์„

์ˆ˜ํ–‰ํ•˜์—ฌ ์›ํ•˜๋Š” ๋™์ž‘์˜ ๊ตฌํ˜„์ด ๊ฐ€๋Šฅํ•จ์„ ํ™•์ธํ•˜์˜€๋‹ค. ๋ฌผ๊ฑด์˜

ํŒŒ์ง€๋ฅผ ์œ„ํ•œ ์„ธ ๊ฐ€์ง€ ๋™์ž‘์— ๋Œ€ํ•œ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ธก์ •ํ•˜๊ณ  ์ธ๊ณต

์‹ ๊ฒฝ๋ง์„ ์ด์šฉํ•˜์—ฌ ํ‰๊ท  90% ์ด์ƒ์˜ ๋†’์€ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๊ฐ€์ง€

๋ฉด์„œ ๋™์ž‘์„ ์ •ํ™•ํ•˜๊ฒŒ ๊ตฌ๋ถ„ํ•  ์ˆ˜ ์žˆ์Œ์„ ํ™•์ธํ•˜์˜€๋‹ค. ์‹ค์‹œ๊ฐ„ ํ™˜

๊ฒฝ์—์„œ, ์ธก์ •๋œ ๊ทผ์ „๋„ ์‹ ํ˜ธ๋ฅผ ์ด์šฉํ•˜์—ฌ ์˜๋„ํ•˜๋Š” ์†๋™์ž‘์„ ์ธ

์‹ํ•˜๊ณ  ์ ์ ˆํ•œ ๋™์ž‘์„ ๊ตฌํ˜„ํ•  ์ˆ˜ ์žˆ์Œ์„ 3D ํ”„๋ฆฐํ„ฐ๋ฅผ ์ด์šฉํ•˜์—ฌ

์ œ์ž‘๋œ ์˜์ˆ˜๋ฅผ ํ™œ์šฉํ•œ ์‹คํ—˜์„ ํ†ตํ•˜์—ฌ ํ™•์ธํ•˜์˜€๋‹ค. ๋‹ค์Œ ๋‹จ๊ณ„

์—ฐ๊ตฌ์—์„œ๋Š” ์ถ”๊ฐ€์ ์œผ๋กœ ๋”์šฑ ๋‹ค์–‘ํ•œ ์†๋™์ž‘ ๊ตฌํ˜„์„ ์œ„ํ•œ ์—ฐ๊ตฌ

๋ฅผ ์ˆ˜ํ–‰ํ•  ์˜ˆ์ •์ด๋ฉฐ, ๋”ฅ๋Ÿฌ๋‹ ๊ธฐ๋ฐ˜์˜ ๋‹ค์–‘ํ•œ ๋ถ„๋ฅ˜๊ธฐ์— ๋Œ€ํ•œ ํ‰๊ฐ€

๋ฅผ ํ†ตํ•˜์—ฌ ๋ถ„๋ฅ˜ ์ •ํ™•๋„๋ฅผ ๋†’์ผ ์ˆ˜ ์žˆ๋„๋ก ํ•  ๊ฒƒ์ด๋‹ค.

ACKNOWLEDGEMENT

๋ณธ ์—ฐ๊ตฌ๋Š” ๊ธˆ์˜ค๊ณต๊ณผ๋Œ€ํ•™๊ต ํ•™์ˆ ์—ฐ๊ตฌ๋น„๋กœ ์ง€์›๋˜์—ˆ์Œ(No.

2018-104-122).

REFERENCES

1. Saridis, G. N. and Gootee, T. P., โ€œEMG Pattern Analysis and

Classification for a Prosthetic Arm,โ€ IEEE Transactions on

Biomedical Engineering, Vol. BME-29, No. 6, pp. 403-412,

1982.

2. Matrone, G. C., Cipriani, C., Secco, E. L., Magenes, G., and

Carrozza, M. C., โ€œPrincipal Components Analysis Based Control

of a Multi-DOF Underactuated Prosthetic Hand,โ€ Journal of

Neuroengineering and Rehabilitation, Vol. 7, No. 1, pp. 1-13,

2010.

3. Ma, J., Thakor, N. V., and Matsuno, F., โ€œHand and Wrist

Movement Control of Myoelectric Prosthesis based on Synergy,โ€

IEEE Transactions on Human-Machine Systems, Vol. 45, No. 1,

pp. 74-83, 2014.

4. Guo, W., Sheng, X., Liu, H., and Zhu, X., โ€œToward an Enhanced

Human-Machine Interface for Upper-Limb Prosthesis Control

with Combined EMG and NIRS Signals,โ€ IEEE Transactions on

Human-Machine Systems, Vol. 47, No. 4, pp. 564-575, 2017.

5. Lee, S., Kim, M. O., Kang, T., Park, J., and Choi, Y., โ€œKnit Band

Sensor for Myoelectric Control of Surface EMG-Based

Prosthetic Hand,โ€ IEEE Sensors Journal, Vol. 18, No. 20, pp.

8578-8586, 2018.

6. Lee, S., Choi, Y., Yang, S., Hong, G. Y., and Choi, Y.,

โ€œDevelopment of Multi-DoFs Prosthetic Forearm based on EMG

Pattern Recognition and Classification,โ€ The Journal of Korea

Robotics Society, Vol. 14, No. 3, pp. 228-235, 2019.

7. Lee, G. H., Kwon, H. C., and Kim, K. H., โ€œImprovement of an

Underactuated Prosthetic Hand based on Grasp Performance

Evaluation,โ€ Transactions of the Korean Society of Mechanical

Engineers, A, Vol. 40, No. 10, pp. 843-849, 2016.

8. Lim, H. S., Kwon, H. C., and Kim, K. H., โ€œDevelopment of a

Haptic System for Grasp Force Control of Underactuated

Prosthetics Hands,โ€ Transactions of the Korean Society of

Mechanical Engineers, A, Vol. 41, No. 5, pp. 415-420, 2017.

9. Bae, J. H., Jung, S. Y., Kim, S., Mun, M., and Ko, C. Y.,

โ€œDevelopment of Sensory Feedback System for Myoelectric

Prosthetic Hand,โ€ Journal of the Korean Society for Precision

Engineering, Vol. 32, No. 10, pp. 851-856, 2015.

Fig. 11 Snapshots of hand motions and realization with prosthetic

robot hand

Page 7: Design of Prosthetic Robot Hand and Electromyography-Based

ํ•œ๊ตญ์ •๋ฐ€๊ณตํ•™ํšŒ์ง€ ์ œ 37๊ถŒ ์ œ 5ํ˜ธ May 2020 / 345

10. Noh, J., Cho, W., and Kim, J. H., โ€œDesign of a Regression

Model for Four Grasping Patterns and Three Grip Force

Intensities of a Myoelectric Prosthetic Hand,โ€ Journal of the

Korean Society for Precision Engineering, Vol. 35, No. 8, pp.

809-816, 2018.

11. Kim, D. E., Yu, J. H., and Sim, K. B., โ€œEEG Feature

Classification for Precise Motion Control of Artificial Hand,โ€

Journal of Korean Institute of Intelligent Systems, Vol. 25, No. 1,

pp. 29-34, 2015.

12. Jo, Y. U. and Oh, D. C., โ€œStudy on the EMG-Based Grasp

Gesture Classification Using Deep Learning and Application to

Active Prosthetics,โ€ Journal of Institute of Control, Robotics and

Systems, Vol. 25, No. 3, pp. 229-234, 2019.

Ho Myoung Jang

M.Sc. candidate in the Department of

Mechanical Design Engineering, Kumoh

National Institute of Technology. His

research interest is design and control of

smart rehabilitation systems with machine

learning algorithm.

E-mail: [email protected]

Jung Woo Sohn

Professor in the Department of Mechanical

Design Engineering, Kumoh National Insti-

tute of Technology. His research interests are

smart systems control, robotics, and applica-

tions of artificial intelligence.

E-mail: [email protected]