[1]王杜,桂凯,曹红升,等.基于肌电控制的辅助型机器人外骨骼[J].燕山大学学报,2018,42(3):219-224.[doi:10.3969/j.issn.1007-791X.2018.03.005]
 WANG Du,GUI Kai,CAO Hongsheng,et al.An assistive robotic exoskeleton based on EMG control[J].Journal of YanShan University,2018,42(3):219-224.[doi:10.3969/j.issn.1007-791X.2018.03.005]
点击复制

基于肌电控制的辅助型机器人外骨骼
分享到:

《燕山大学学报》[ISSN:1007-791X/CN:13-1219/N]

卷:
42
期数:
2018年第3期
页码:
219-224
栏目:
机械工程
出版日期:
2018-05-31

文章信息/Info

Title:
An assistive robotic exoskeleton based on EMG control
文章编号:
1007-791X(2018)03-0219-06
作者:
王杜桂凯曹红升张定国*
上海交通大学 机械与动力工程学院,上海 201100
Author(s):
 WANG DuGUI KaiCAO HongshengZHANG Dingguo
School of Mechanical Engineering,Shanghai Jiao Tong University,Shanghai 201100,China
关键词:
臂丛神经损伤外骨骼系统表面肌电信号线性判别分析在线识别
Keywords:
brachial plexus injuryexoskeletonsurface electromyographylinear discriminant analysisonline recognition
分类号:
TP242
DOI:
10.3969/j.issn.1007-791X.2018.03.005
文献标志码:
A
摘要:
针对臂丛神经损伤患者上肢功能部分或完全丧失的问题,本文设计了一款基于肌电控制的辅助型机器人外骨骼。该外骨骼具有2个自由度,受试者穿戴上之后能完成5个上肢动作。采用腱鞘驱动方式设计外骨骼结构,减轻受试者上肢使用负担。通过提取受试者健侧手臂产生的肌电信号,使用基于贝叶斯决策的线性判别分析方法解码受试者运动意图,实现利用健侧上肢运动带动患侧运动的目的。而本研究最终目标则是通过解码患侧肌电信号控制患侧上肢运动,实现受试者在不需要任何帮助的前提下重建患侧上肢运动功能。系统对肌电信号的处理、识别均为在线进行,整个过程耗时约57 ms,完全符合外骨骼使用过程中的延时要求。对4名健康受试者进行实验,结果表明5类动作的平均在线识别率均达到95%以上,且受试者能够顺畅使用所设计外骨骼系统。
Abstract:
An assistive robotic exoskeleton based on EMG control is developed for patients with brachial plexus injury who have lost their upper limb function partially or completely in this paper.The wearable exoskeleton has 2 degrees of freedom and the patient can perform 5 types of upper limb movements.The exoskeleton structure is designed by tendon-driven mechanism which can reduce the burden on subjects′ upper limb.The exoskeleton can help the subject to ipsilateral upper limb by extracting subjects′ contralateral surface electromyography and decoding his motivation with linear discriminant analysis method based on Bayesian decision in order to rebuild subject′s ipsilateral upper limb function without any help by decoding his ipsilateral surface electromyography.The system′s processing and recognition of EMG signal is performed online,and the entire process takes about 57 ms,which satisfy the delay requirements in the use of exoskeleton.Experiments are conducted on four healthy subjects. The results show that the average online recognition rate of the five types of movements is over 95%.The subjects are able to use the exoskeleton system smoothly.

备注/Memo

备注/Memo:
 收稿日期:2018-01-05      责任编辑:温茂森
基金项目:国家自然科学基金资助项目(51475292)
作者简介:王杜(1993-),男,湖北黄冈人,硕士研究生,主要研究方向为肌电控制和康复机器人;*通信作者:张定国(1978-),男,吉林辉南人,博士,副教授,博士生导师,主要研究方向为生机电一体化、康复机器人、仿生控制,Email:dgzhang@sjtu.edu.cn
更新日期/Last Update: 2018-07-16