基于滤波器组典型相关分析的SSVEP信号分类方法  

Classification method of SSVEP signals based on filter bank canonical correlation analysis

作　　者：马洪远 张学军[1,2] MA Hongyuan;ZHANG Xuejun(College of Electronic and Optical Engineering,College of Flexible Electronics(Future Technology),Nanjing University of Posts and Telecommunications,Nanjing 210023,China;Nation and Local Joint Engineering Laboratory of Radio Frequency Integration and Micro-Assembly Technologies,Nanjing University of Posts and Telecommunications,Nanjing 210023,China)

机构地区：[1]南京邮电大学电子与光学工程学院,柔性电子(未来技术)学院,南京210023 [2]南京邮电大学射频集成与微组装技术国家地方联合工程实验室,南京210023

出　　处：《智能计算机与应用》2024年第7期29-36,共8页Intelligent Computer and Applications

基　　金：国家自然科学基金(61977039)。

摘　　要：基于稳态视觉诱发电位(SSVEP)的脑机接口系统存在鲁棒性不足和短时分类准确率较低的问题,本文提出了一种基于滤波器组典型相关分析的SSVEP信号分类方法(FBCCA-SVM)。通过结合性能较为优秀的滤波器组典型相关分析(FBCCA)和鲁棒性强的支持向量机(SVM)的优点,使用SVM分类器替代FBCCA中的MAX分类器对FBCCA提取到的特征分类,在较短时间的刺激下在多目标分类中达到了较高的准确率和信息传输速率(ITR)。通过对8名受试者的实验结果表明,2 s的刺激时间下平均准确率为93.91%,信息传输速率为78.63 bit/min,与典型相关分析方法相比,平均准确率和信息传输速率分别提高了14.32%和25 bit/min;与滤波器组典型相关分析方法相比,分别提高了6.36%和9.66 bit/min;相比于主流识别算法,该方法性能得到明显提高,同时增强了目标分类的鲁棒性,为后续脑机接口系统的实际应用提供了实验基础。Brain-computer interface systems based on steady-state visual evoked potential(SSVEP)have problems of poor robustness and low short-term classification accuracy.In this paper,a classification method based on FBCCA-SVM is proposed.By combining the advantages of the filter bank canonical correlation analysis(FBCCA)and the robust support vector machine(SVM),the SVM classifier is used to replace the MAX classifier in the FBCCA algorithm to classify the features extracted by FBCCA,and the accuracy and information transmission rate(ITR)are higher in the multi-objective classification under the stimulation of a short time.The results of eight subjects show that the average accuracy is 93.91%and ITR is 78.63 bit/min for 2 seconds.Compared with CCA,the average accuracy and ITR are increased by 14.32%and 25 bit/min,respectively.Compared with FBCCA,the increase is 6.36%and 9.66 bit/min,respectively.In comparison with the mainstream recognition algorithm,the performance of this method is significantly improved,and the robustness of target classification is enhanced,which provides an experimental basis for the subsequent practical application of brain-computer interface system.

关 键 词：稳态视觉诱发电位 脑机接口 滤波器组典型相关分析 支持向量机 目标分类 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] R318[自动化与计算机技术—控制科学与工程]