SIA OpenIR  > 机器人学研究室
基于自更新混合分类模型的肌电运动识别方法
Alternative TitleAn EMG-motion Recognition Method With Self-update Hybrid Classification Model
丁其川1; 赵新刚1; 李自由1,2; 韩建达1,3
Department机器人学研究室
Source Publication自动化学报
ISSN0254-4156
2019
Volume45Issue:8Pages:1464-1474
Indexed ByEI ; CSCD
EI Accession number20194307576164
CSCD IDCSCD:6552935
Contribution Rank1
Funding Organization国家高技术研究发展计划(863计划) (2015AA042301)资助
Keyword表面肌电 动作识别 模式分类 在线更新 肌肉疲劳
Abstract

传统基于肌电的运动识别方法多是利用训练后的固定参数模型,分类已预先定义的有限个目标动作,但对肌肉疲劳导致的肌电变化,以及未定义的外部动作等干扰因素无能为力.针对这一问题,提出一种自更新混合分类模型(Self-update hybrid classification model,SUHC),该模型融合了用于排除外部动作干扰的一类支持向量机,以及用于分类目标动作数据的多类线性判别算法,并引入自更新机制以对抗肌电时变性干扰.通过手部动作识别实验验证提出方法的效果,在肌电大幅变化干扰下,SUHC的目标动作识别精度达到89%,对比传统的支持向量机、多层感知器和核线性判别分析,提高了约18%,并且SUHC具备排除外部动作干扰能力,排除精度高达93%.

Other Abstract

The traditional EMG-based motion recognition methods always classify a limited of pre-defined target motions by using a trained parameter-fixed model. However, those methods have no ability to handle the interference factors, including changes of EMG signals caused by muscle fatigue, and outlier motions undefined beforehand. With respect to the problem, a self-update hybrid classification model (SUHC) was proposed. In the SUHC, the one-class SVM (Support vector machine) that can reject outlier motions is combined to a multi-class LDA (Linear discriminant analysis) that is used to recognize target motions; furthermore, a self-update mechanism is introduced to reduce the inuence caused by EMG changes. The performance of the proposed method was verified by the experiments of EMG-based hand motion recognition. Under the interference that EMG signals varied greatly, the recognition accuracy on target motions of SUHC is about 89 %, which is 18% higher than that of the normal SVM, MLP (Multiple layer perceptron) and KLDA (Kernel LDA); moreover, the SUHC has the ability to reject outlier motions with a 93% of rejection accuracy.

Language中文
Citation statistics
Document Type期刊论文
Identifierhttp://ir.sia.cn/handle/173321/21580
Collection机器人学研究室
Corresponding Author丁其川
Affiliation1.中国科学院沈阳自动化研究所机器人学国家重点实验室
2.中国科学院大学
3.南开大学计算机与控制工程学院
Recommended Citation
GB/T 7714
丁其川,赵新刚,李自由,等. 基于自更新混合分类模型的肌电运动识别方法[J]. 自动化学报,2019,45(8):1464-1474.
APA 丁其川,赵新刚,李自由,&韩建达.(2019).基于自更新混合分类模型的肌电运动识别方法.自动化学报,45(8),1464-1474.
MLA 丁其川,et al."基于自更新混合分类模型的肌电运动识别方法".自动化学报 45.8(2019):1464-1474.
Files in This Item: Download All
File Name/Size DocType Version Access License
基于自更新混合分类模型的肌电运动识别方法(2199KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[丁其川]'s Articles
[赵新刚]'s Articles
[李自由]'s Articles
Baidu academic
Similar articles in Baidu academic
[丁其川]'s Articles
[赵新刚]'s Articles
[李自由]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[丁其川]'s Articles
[赵新刚]'s Articles
[李自由]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 基于自更新混合分类模型的肌电运动识别方法.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.