摘要
提出一种基于模糊化符号复杂度的运动想象脑电信号特征提取与识别方法。在脑电信号的复杂度细粒化多符号度量中引入模糊算法,用sigmoid函数模糊化处理,逻辑判断得到模糊化符号复杂度。取细粒化指数n为2,提取模糊化符号复杂度作为特征值,最后利用支持向量机对脑电运动想象任务进行分类识别。实验结果表明,以模糊化符号复杂度为特征的分类方法,对左右手运动想象脑电信号的分类识别率最高达88.67%,优于二值化Lempel-Ziv复杂度算法。
A method of Electroencephalogram (EEG)feature extraction and recognition of motor imagery based on fuzzy symbolic complexity is proposed. Introduce Fuzzy algorithm in the EEG complexity fine-grained and multi- symbol metrics, fuzzy processing with the sigmoid function, and calculate fuzzy symbolic complexity by logical judgment. Select the fine graining index n as 2, extract fuzzy symbolic complexity as a characteristic value, and finally use the Support Vector Machine to classify EEG consciousness task of motor imagery. The experimental result shows that the average classification accuracy of EEG of two hands motor imagery can reach 88.67% to the highest owing to the classification method featured by fuzzy symbolic complexity, which excels the algorithm of binary quan- tification Lempel-Ziv complexity.
出处
《传感技术学报》
CAS
CSCD
北大核心
2013年第5期595-599,共5页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(61172134)
浙江省自然科学基金项目(LY12F03006)
关键词
运动想象
模糊化
脑机接口
特征提取
motor imagery
fuzzy symbolic complexity
brain-computer interface
feature extraction