基于总体经验模态分解的多类特征的运动想象脑电识别方法研究
发布时间：2018-08-06 21:29
【摘要】：人的脑电信号(Electroencephalogram,EEG)复杂且具有非线性及非平稳性的特点使其不易分析处理,其识别效果也依赖于数据集的不同,而表现不稳定.本文中应用的总体经验模态分解(Ensemble empirical mode decomposition,EEMD)是一种具有强自适应性的信号处理方法,其在时频域展现的良好分辨率特别适合脑电识别任务处理.本文提出利用EEMD分解后得到的较具影响能力的固有模态函数(Intrinsic mode functions,IMFs),利用希尔伯特变换提取边际谱(Marginal spectrum,MS)及瞬时能谱(Instantaneous energy spectrum,IES)时频特征,同时通过加窗的方法提取非线性动力学特征近似熵特征,利用线性判别分类器(Linear discriminant analysis,LDA)作为分类器,实验结果得出,对于被试S2和被试S3可达到识别率分别为79.60%和87.77%,实验中9名被试的平均识别率为82.74%,得到平均识别率也高于近期使用相同数据集文献的其他方法.
[Abstract]:The complex, nonlinear and non-stationary characteristics of electroencephalogram (EEG) make it difficult to analyze and process, and its recognition effect depends on the difference of data set, and its performance is unstable. The general empirical mode decomposition (Ensemble empirical mode) method used in this paper is a strong adaptive signal processing method, and its good resolution in time-frequency domain is especially suitable for EEG recognition task processing. In this paper, the time-frequency features of Marginal spectra MS and instantaneous energy spectrum (Instantaneous energy spectra are extracted by using the (Intrinsic mode functionsimfs, which are obtained by EEMD decomposition, and by Hilbert transform. At the same time, the approximate entropy feature of nonlinear dynamics is extracted by adding windows, and the linear discriminant classifier (Linear discriminant analysisLDA is used as the classifier. The experimental results show that, The recognition rates of S2 and S3 were 79.60% and 87.77%, respectively. The average recognition rate of 9 subjects was 82.74, and the average recognition rate was higher than that of other methods using the same data set recently.
【作者单位】： 吉林大学通信工程学院分布式智能信息处理实验室;
【基金】：吉林省科技发展计划自然基金(20150101191JC) 吉林大学研究生创新基金(2016092)资助~~
【分类号】：R338
本文编号：2169045
[Abstract]:The complex, nonlinear and non-stationary characteristics of electroencephalogram (EEG) make it difficult to analyze and process, and its recognition effect depends on the difference of data set, and its performance is unstable. The general empirical mode decomposition (Ensemble empirical mode) method used in this paper is a strong adaptive signal processing method, and its good resolution in time-frequency domain is especially suitable for EEG recognition task processing. In this paper, the time-frequency features of Marginal spectra MS and instantaneous energy spectrum (Instantaneous energy spectra are extracted by using the (Intrinsic mode functionsimfs, which are obtained by EEMD decomposition, and by Hilbert transform. At the same time, the approximate entropy feature of nonlinear dynamics is extracted by adding windows, and the linear discriminant classifier (Linear discriminant analysisLDA is used as the classifier. The experimental results show that, The recognition rates of S2 and S3 were 79.60% and 87.77%, respectively. The average recognition rate of 9 subjects was 82.74, and the average recognition rate was higher than that of other methods using the same data set recently.
【作者单位】： 吉林大学通信工程学院分布式智能信息处理实验室;
【基金】：吉林省科技发展计划自然基金(20150101191JC) 吉林大学研究生创新基金(2016092)资助~~
【分类号】：R338
【相似文献】
相关期刊论文 前1条
1 刘燕;穆振东;;运动想象机制研究[J];科技广场;2011年01期
相关硕士学位论文 前8条
1 王超;基于OpenVIBE的高性能运动想象在线脑—机接口系统研究[D];华南理工大学;2016年
2 马秀秀;基于运动想象的异步脑—机接口算法研究[D];哈尔滨工业大学;2016年
3 王延达;基于运动想象的脑电信号识别算法研究[D];南京航空航天大学;2016年
4 张f欢;运动想象脑电信号的分析研究[D];上海交通大学;2015年
5 姚家扬;运动想象脑电处理及其模式识别方法研究[D];杭州电子科技大学;2016年
6 蔡慧;BCI中的运动想象脑电信号处理方法研究[D];杭州电子科技大学;2016年
7 张焕;四类运动想象任务的脑电信号识别算法研究[D];山西大学;2016年
8 易振华;运动想象脑电信号的分析与处理方法研究[D];北京交通大学;2008年
，本文编号：2169045
本文链接：https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2169045.html
