基于多特征和BP神经网络的脑-机接口研究

发布时间：2018-04-01 22:03

本文选题：多特征　切入点：BP神经网络　出处：《电子技术应用》2017年09期

【摘要】：研究了一种基于运动想象识别的脑-机接口(BCI)系统,通过提取想象过程中的脑电信号(EEG)中Alpha波特征,采用多特征分类的方法,以提高脑-机接口系统运动想象识别的正确率。针对脑电信号单特征分类精确度低、耗时长等缺点,采用自回归模型法、统计特征提取和频域分析的方法对Alpha波提取多个特征值,利用BP神经网络进行分类,对运动想象进行识别。通过实验验证了其识别率较高,取得了预期的效果,证明了多特征融合结合BP神经网络运用于脑机接口系统的可行性。
[Abstract]:In this paper, a brain-computer interface (BCI) system based on motion imagination recognition is studied. By extracting the features of Alpha wave in the process of imagining, the method of multi-feature classification is adopted. In order to improve the correct rate of motion imagination recognition in brain-computer interface system, the autoregressive model method is adopted to solve the disadvantages of low accuracy and long time consuming in single feature classification of EEG signals. The methods of statistical feature extraction and frequency domain analysis are used to extract multiple eigenvalues of Alpha wave, and BP neural network is used to classify and recognize the motion imagination. The experimental results show that the recognition rate is high and the expected results are obtained. The feasibility of applying multi-feature fusion and BP neural network to the brain-computer interface system is proved.
【作者单位】：吉林大学仪器科学与电气工程学院;
【分类号】：R318;TN911.7;TP183

【相似文献】