当前位置:主页 > 医学论文 > 生物医学论文 >

基于肌电信号的下肢关节连续运动预测

发布时间:2018-02-11 07:38

  本文关键词: 连续运动预测 肌电信号 支持向量机 运动模式 下肢运动 出处:《华中科技大学学报(自然科学版)》2017年10期  论文类型:期刊论文


【摘要】:为研究不同运动模式下基于肌电信号的下肢多关节连续运动预测,通过支持向量机对肌电-运动的映射关系进行训练,实现对下肢髋、膝和踝3个关节矢状面内的连续运动预测.由10位健康受试者的运动预测和统计分析可知:在适速行走过程中,髋、膝和踝关节的关节角度预测均方根误差分别9.36°,10.82°和6.87°;在不同运动模式下,关节的运动预测值与测量值之间均表现出一定的相关性,其中,膝和髋关节的预测值与测量值之间相关系数均大于0.72,表现出比较明显的相关性.实验结果表明:基于肌电信号进行下肢多关节连续运动预测,尤其是在适速行走时对膝和髋关节的运动预测是可行的.
[Abstract]:In order to study the continuous motion prediction of multiple joints in lower extremities based on EMG signals in different motion modes, the mapping relationship between electromyography and motion was trained by support vector machine (SVM), and the hip joint of lower extremity was realized. Continuous motion prediction in sagittal plane of knee and ankle. The root mean square error of knee and ankle joint angle prediction was 9.36 掳10.82 掳and 6.87 掳, respectively. The correlation coefficient between the predicted value and the measured value of knee and hip joint is more than 0.72, which shows obvious correlation. The experimental results show that the multiple joint continuous motion prediction of lower extremities is based on EMG signal. Especially, it is feasible to predict the motion of knee and hip joint when walking at proper speed.
【作者单位】: 华中科技大学机械科学与工程学院;
【基金】:中央高校基本科研业务费专项资金资助项目(2016YXMS274) 国家自然科学基金资助项目(91648203) 科技部重点研发计划资助项目(2016YFE0113600)
【分类号】:R318.04


本文编号:1502528

资料下载
论文发表

本文链接:https://www.wllwen.com/yixuelunwen/swyx/1502528.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户f56e5***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com