多通道三维视觉指导运动想象脑电信号特征选择算法

针对基于三维视觉指导的运动想象脑机接口多通道冗余信息较多、分类准确率差的问题，提出了一种基于小波包分解（WPD）—共空间滤波（CSP）—自适应差分进化（ADE）的模式脑电信号特征提取与选择分类方法。首先，对采集的多通道运动想象脑电信号进行WPD变化，划分出精细的子频带；然后，分别将WPD变换后的每个子空间作为CSP的输入，得到对应的特征向量；最后，使用ADE算法对特征向量进行选择，选择出用于分类的最佳特征子集。采用WPD-CSP-ADE模式进行特征提取与选择，较经典的WPD-CSP方法在分类正确率、特征个数方面有着更好的表现。同时，所提算法分类性能明显优于遗传算法、粒子群算法。实验结果表明，WPD-CSP-ADE方法能够有效地提高分类正确率，同时减少了用于分类的特征个数。

Feature selection algorithm for 3D visual guidance multi-channel motor imagery

Concern the problem that multi-channel motor imagery of BCI based on 3D visual guidance with more redundancy information and poor classification accuracy, this paper proposed a pattern classification method based on wavelet packet decomposition(WPD) -common spatial pattern(CSP) -ADE for feature extraction of EEG. Firstly, this algorithm used WPD to divide the multi-channel motion imagery EEG signals into fine sub-bands. Secondly, it used CSP to obtain the eigenvectors corresponding to each subspace of WPD transformation. Finally, it selected the feature vectors through the ADE algorithm to obtain the best feature subsets for classification. Using WPD-CSP-ADE mode for feature extraction and selection, it has better performance in classification accuracy and number of features than the classic WPD-CSP method. At the same time, the classification performance of the proposed algorithm is significantly better than the genetic algorithm and particle swarm optimization algorithm. The experiments show that the WPD-CSP-ADE method can effectively improve the classification accuracy and reduce the number of features used for classification.