基于小波分析和SVM的P300脑电信号识别算法研究

摘要

为了满足瘫痪人士和虚拟现实的需求,提出基于小波分析和SVM的P300脑电信号处理算法研究,并通过实验数据论证算法的可行性.本算法首先使用工频陷波器和小波分析去噪,然后使用小波分解和teager能量算子分别提取时域特征量和能量特征量,并基于SVM判断特征量是否含有P300脑电信号.实验数据表明,本算法比单一特征量判别算法有较好的判别精度,符合需求标准.%In order to meet the requirements of the paralyzed clinical and virtual reality,proposed based on wavelet analysis and SVM of P300 EEG signals processing algorithm research,and through the experimental data prove the feasibility of this algorithm.The algorithm first makes power frequency filter and wavelet denoising,and then use the wavelet decomposition and teager energy operator respectively extract the time domain characteristics and energy characteristics,and based on the SVM judge whether characteristics contain P300 EEG signals.The experimental data show that this algorithm is better than the single feature discriminant algorithm,accord with the standard requirements