基于窗函数法的低频肌电信号异常分类仿真  

Simulation of Abnormal Classification of Low Frequency EMG Signals Based on Window Function Method

作　　者：杨航 康晶晶[2] YANG Hang;KANG Jing-jing(School of Electronic Information Engineering,Henan Institute of Technology,Xinxiang Henan 453003,China;College of Information,Shanxi Agricultural University,Taigu Shanxi 030800,China)

机构地区：[1]河南工学院电子信息工程学院,河南新乡453003 [2]山西农业大学信息学院,山西太谷030800

出　　处：《计算机仿真》2021年第11期244-248,共5页Computer Simulation

摘　　要：为准确实现对异常低频肌电信号的分析和处理,基于窗函数法设计了一种低频肌电信号异常分类方法。首先通过加窗过程修正低频肌电信号及其相关参数,并分解重构每一周期的低频肌电信号,去除其中的噪声干扰并提取完成修正的信号时域与频域特征。识别异常的低频肌电信号,融合时域与频域特征并表征低频肌电信号,再通过线性分类器实现对低频肌电信号的异常分类。仿真结果表明,上述方法对低频肌电信号异常分类的平均准确率高,可精准识别异常的低频肌电信号,且其在不同噪声影响下受噪声影响程度较小,分类效果更好。This paper designd an abnormal classification method of low-frequency EMG signals based on window function method in order to accurately analyze and process abnormal low-frequency EMG signals. First of all, the low-frequency EMG signal and related parameters were modified by the windowing process. Meanwhile, all the low-frequency EMG signals in each cycle were decomposed and reconstructed. Then, the noise in the signal was removed and the time-domain and frequency-domain characteristics of the modified signal were extracted. Secondly, abnormal low-frequency EMG signals were identified, and time-domain and frequency-domain features were combined to characterize low-frequency EMG signals. Finally, based on the linear classifier, the anomaly classification of low-frequency EMG signals was achieved. The simulation results show that this method has high average accuracy, excellent classification effect and small interference.

关 键 词：窗函数法 低频肌电信号 异常分类 线性分类器 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]