摘要
为了实现不同运动模式下膝关节连续运动的有效估计,提出一种基于核主成分分析(KPCA)的下肢膝关节连续运动估计方法。首先,融合多维表面肌电信号时域特征获取不同运动模式下较为全面的运动信息;其次,采用KPCA方法进行肌电特征降维,获取与该类运动模式最为相关的主成分向量,并基于反向传播神经网络实现不同运动模式下膝关节连续运动的有效估计;最后,对5个实验对象的4种运动模式进行实验验证。结果表明该方法不仅可有效估计不同运动模式下膝关节连续运动角度,相对于PCA算法估计精度也有明显提高。
A kernel principal component analysis(KPCA) based method is proposed for the effective estimation of multi-mode continuous motion of knee joint.The time domain features of multi-dimensional surface electromyography signal are fused to obtain the comprehensive motion information in different motion modes.Then,KPCA method is used to reduce the dimensionality of EMG features and obtain the most relevant principal component vectors,and the effective estimation of multi-mode continuous motion of knee joint is realized through the combination with back propagation neural network.Experimental verification is carried out on 4 motion modes of 5 subjects.The results show that the proposed method can not only effectively estimate the multi-mode continuous motion angles of knee joint,but also significantly improve the estimation accuracy as compared with PCA algorithm.
作者
张建华
王豪
李克祥
王唱
ZHANG Jianhua;WANG Hao;LI Kexiang;WANG Chang(School of Mechanical Engineering,Hebei University of Technology,Tianjin 300401,China;State Key Laboratory of Reliability and Intelligence of Electrical Equipment,Hebei University of Technology,Tianjin 300401,China)
出处
《中国医学物理学杂志》
CSCD
2023年第6期742-749,共8页
Chinese Journal of Medical Physics
基金
河北省自然科学基金(F20202051)
河北省博士后科研项目(B2022003016)
省部共建电工装备可靠性与智能化国家重点实验室人才培育项目(EERIPD2021011)
天津市杰出青年科学基金(19JCJQJC61600)。
关键词
表面肌电信号
核主成分分析
时域特征
连续运动估计
surface electromyography signal
kernel principal component analysis
time domain feature
continuous motion estimation
