基于自适应VMD的sEMG信号识别研究  

Research on sEMG Signal Recognition Based on Adaptive VMD

作　　者：胡家铭 曾庆军[2] 韩春伟 周成诚 HU Jiaming;ZENG Qingjun;HAN Chunwei;ZHOU Chengcheng(College of Computer Science,Jiangsu University of Science and Technology,Zhenjiang Jiangsu 212100,China;College of Automation,Jiangsu University of Science and Technology,Zhenjiang Jiangsu 212100,China)

机构地区：[1]江苏科技大学计算机学院,江苏镇江212100 [2]江苏科技大学自动化学院,江苏镇江212100

出　　处：《电子器件》2025年第1期25-30,共6页Chinese Journal of Electron Devices

基　　金：国家自然科学基金项目(11574120);江苏省产业前瞻与共性关键技术项目(BE2018103)。

摘　　要：针对表面肌电(sEMG)信号噪声导致基于sEMG信号的手势识别准确率不高的问题,提出了一种基于自适应变分模态分解(AVMD)的sEMG信号手势识别算法。首先,采用AVMD算法和改进小波阈值对sEMG信号进行降噪;然后提取sEMG信号的均值和模糊熵作为特征值;最后,采用支持向量机(SVM)进行手势识别。实验结果表明,基于AVMD的sEMG信号手势识别方法降噪性能指标高于其他方法,手势识别准确率达到97.5%,并能在手部康复机器人主从训练系统中准确实时识别出对应的手势动作。To solve the problem of low accuracy of gesture recognition based on sEMG signal due to surface electromyography(sEMG)signal noise,a gesture recognition algorithm based on adaptive variational mode decomposition(AVMD)of sEMG signal is proposed.Firstly,the sEMG signal is denoised by using AVMD algorithm and improved wavelet threshold.Then,the mean value and fuzzy entropy of sEMG signal are extracted as eigenvalues.Finally,support vector machine(SVM)is used for gesture recognition.The experimental re-sults show that the noise reduction performance index of the sEMG signal gesture recognition method based on AVMD is higher than that of other methods,the gesture recognition accuracy reaches 97.5%,and the corresponding gestures can be accurately recognized in real time in the master-slave training system of the hand rehabilitation robot.

关 键 词：手势识别 表面肌电信号 自适应变分模态分解 信号降噪 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]