运动疲劳过程中脑电信号特征提取仿真  

Sports Fatigue in the Electrical Signal Feature Extraction Simulation

作　　者：高明信[1] 

机构地区：[1]郑州大学西亚斯国际学院,河南新郑451150

出　　处：《计算机仿真》2017年第5期277-280,共4页Computer Simulation

基　　金：基金项目:2015年河南省教育技术装备和实践教育研究(GZS084)

摘　　要：对运动疲劳过程中脑电信号特征进行准确提取,可以为运动疲劳损伤的治疗提供科学依据。对运动疲劳过程中不同层次脑电信号中的脑电波进行分解是进行脑电信号特征提取的基础,而传统方法利用可预测性的选取嵌入维数方法计算脑电信号序列的嵌入维数,利用相空间重构对整体信号特征提取,但是不能对不同层次脑电波进行分解,导致脑电信号特征提取精度差。提出基于模糊熵的运动疲劳过程中脑电信号特征提取方法。上述方法先融合于小波变换理论,将运动疲劳过程中不同层次脑电信号中的脑电波进行合理分解,计算不同层次脑电节律频带中小波系数的能量均值与均值差,将能量均值与均值差作为特征向量,构建FISHER线性分类器对运动疲劳中的意识疲劳信号分类。仿真结果表明,所提方法可以有效地完成对运动疲劳过程中脑电信号特征提取。A feature extraction method of electroencephalogram signal during sports fatigue is proposed based on fuzzy entropy. The electroencephalogram signal during sports fatigue is firstly decomposed into different levels reason- ably based on wavelet transform theory. Then the energy mean value and mean difference of wavelet coefficient in fre- quency band of electroencephalogram rhythm with different levels are calculated. Finally, the FISHER linear classifi- er is built to classify consciousness fatigue signal during sports fatigue by using energy mean value and mean differ- ence as feature vector. The simulation results show that the method mentioned above can complete the feature extrac- tion of electroencephalogram signal effectively.

关 键 词：脑电信号 特征提取 运动疲劳 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]