JAIST Repository >
School of Information Science >
Grants-in-aid for Scientific Research Papers >
FY 2010 >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10119/9797

Title: 海馬型ニューラルネットに基づいた筋電義手のための動作推定システムの開発
Other Titles: Motion Estimation System for Myoelectric Prosthetic Hand Based on a Hippocampus Neural Network
Authors: 末光, 厚夫
Authors(alternative): Suemitsu, Atsuo
Keywords: 医療・福祉
Issue Date: 10-Jun-2011
Abstract: 本研究では,使い勝手の良い筋電義手の実現のために,申請者らが提案した海馬型ニューラルネットを基に,表面筋電位(EMG)信号から使用者の意図した手・腕の動きを推定するシステムを構築した.構築したシステムは,未学習のEMG 信号から6動作(手首屈曲,手首伸展,握る,開く,前腕回内,前腕回外)を平均約95%の精度で識別可能であるだけでなく,大量の学習データは不要,センサの冗長性を許容,事前トレーニングは不要という実用的な特徴を持つ. : In this study, we constructed a motion estimation system for practical myoelectric prosthetic hands based on a hippocampus neural network we have previously proposed. The proposed system can classify six hand motions (wrist flexion, wrist extension, grasping, opening up, wrist supination, and wrist pronation) with an average of about 95% accuracy by using surface electromyogram signals. In addition, it does not require the large number of training data samples, the user to position sensors on optimal locations, and the user to be trained in advance.
Description: 若手研究(B)
Language: jpn
URI: http://hdl.handle.net/10119/9797
Appears in Collections:2010年度 (FY 2010)

Files in This Item:

File Description SizeFormat
21700576seika.pdf335KbAdobe PDFView/Open

All items in DSpace are protected by copyright, with all rights reserved.


Contact : Library Information Section, Japan Advanced Institute of Science and Technology