摘要
表面肌电(surface electromyogram,s EMG)信号的去噪处理和特征提取的效果好坏直接关系到识别的准确率。以获得较高的识别准确率为目标,对肌电信号的去噪处理和特征提取展开研究。先对表面肌电信号进行小波阈值去噪;再分别运用时域、频域和时频分析对去噪后的信号进行特征提取;最后利用BP神经网络对肌电信号进行分类。实验结果较好地实现了对肌电信号的分类,分类识别率为97%±2%。
The effect of denoising and feature extraction of sEMG is directly realated to the accuracy of recognition. To achieve a higher accuracy of recognition, the research on the optimization of denoising and feature extraction is needed.Firstly, the surface EMG signal is denoised by wavelet threshold. Then the time domain, frequency domain and time-frequency analysis are used to extract the features of the denoised signal;Finally, BP neural network is used to classify sEMG signals. The experimental results show that the classification of EMG signals is well realized, and the recognition rate is 97%±2%.
作者
顾兴龙
宋天赐
陈文涛
毛嘉元
GU Xinglong;SONG Tianci;CHEN Wentao;MAO Jiayuan(Engineering Technology Training Center,Civil Aviation Flight University of China,Guanghan 618307,China;College of Automation,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)
出处
《机械工程师》
2022年第5期17-19,共3页
Mechanical Engineer
基金
中国民用航空飞行学院研究生教学建设项目(XKJ2019-3)
中国民用航空飞行学院青年基金(Q2018-34)。
关键词
表面肌电信号
特征提取
小波包变换
神经网络
surface electromyogram signal(sEMG)
feature extraction
wavelet packet transform
neural network
