一种基于PSO-CSP-SVM的运动想象脑电信号特征提取及分类算法

来源期刊:中南大学学报(自然科学版)2020年第10期

论文作者:刘宝 蔡梦迪 薄迎春 ZHANG Xin(张欣)

文章页码:2855 - 2867

关键词:运动想象;粒子群优化;共空间模式;支持向量机

Key words:motor imagery; particle swarm optimization; common spatial pattern; support vector machine

摘    要:针对因脑电信号存在个体差异性而造成多类运动想象脑电信号特征提取困难和识别正确率较低的问题,提出一种基于PSO-CSP-SVM的运动想象脑电信号特征提取及分类算法。首先,利用粒子群优化(PSO)算法优化得到不同个体脑电信号的最佳时间段和频段;然后,基于优化时频段的脑电信号,利用“一对多”共空间模式(OVR-CSP)算法进行特征提取,将特征向量输入到“一对一”支持向量机(OVO-SVM)中实现分类,并且将分类错误率作为PSO算法的适应度函数值;最后,采用BCI2005desc_Ⅲa数据集验证该算法的分类效果。研究结果表明:相比基于固定时频段脑电信号得到的分类结果以及其他文献中算法的分类结果,该算法的平均分类准确率有较大提高,达87.65%,证明该算法能够有效提取脑电特征,并且具有较好的运动想象脑电信号识别效果。

Abstract: Aiming at the problems of difficulty in feature extraction and the low recognition accuracy for multiple types of motor imagery EEG signals due to individual differences in EEG signals, a feature extraction and classification algorithm based on PSO-CSP-SVM for motor imagery EEG signals was proposed. Firstly, particle swarm optimization(PSO) algorithm was used to optimize the optimal time and frequency band of EEG signals for different individuals. Then, the EEG signals based on the optimized time and frequency band were extracted feature by using the "one versus rest" common spatial patterns(OVR-CSP) algorithm. Finally, the feature vectors were input into the "one versus one" support vector machine(OVO-SVM) to achieve classification, and the classification error rate was taken as the fitness function value of the PSO algorithm. In the experiment, BCI2005desc_Ⅲa data set was employed to verify the classification effect of the algorithm. The results show that compared with the classification results based on fixed time and frequency band EEG signals and the classification results of other algorithms in the literature, the average classification accuracy of the proposed algorithm is greatly improved, which reaches 87.65%. It is proved that the proposed algorithm can extract EEG features effectively and has a better recognition effect of EEG signals of motor imagery.

相关论文

  • 暂无!

相关知识点

  • 暂无!

有色金属在线官网  |   会议  |   在线投稿  |   购买纸书  |   科技图书馆

中南大学出版社 技术支持 版权声明   电话:0731-88830515 88830516   传真:0731-88710482   Email:administrator@cnnmol.com

互联网出版许可证:(署)网出证(京)字第342号   京ICP备17050991号-6      京公网安备11010802042557号