多重分形分析在肌电信号识别中的应用  

The Application of Multifractal in SEMG Pattern Recognition

作　　者：宋玲玲[1] 奚日辉[1] 吴小丹[1] 

机构地区：[1]上海大学通信与信息工程学院,上海200072

出　　处：《微计算机信息》2008年第13期272-273,260,共3页Control & Automation

摘　　要：针对表面肌电信号的非线性特征,目前的分析方法有非线性动力学的复杂度及分形维数等,但是单重分形只能反映信号的整体特征,缺乏对局部奇异性的刻画。本文采用多重分形来分析表面肌电信号,结合灰色关联度对四种前臂动作进行分类识别,取得了较好的识别效果。结果表明:广义维数谱可以作为一种新的肌电信号特征来描述肌电信号的复杂性,为表面肌电信号的识别提供了一种新的思路。To overcome the obstacle of Surface Electromyography （SEMG） nonlinear characteristic, many methods had been proposed, such as complexity, fractal dimension of nonlinear dynamics. However, single fractal shows the global feature of signal, lacks of the information of the local singularity. Muhifractal was adopted in the thesis, and grey cOrrelation analysis as a new method is successfully put forward to classifying and recognizing of different movement patterns. Experiments show that a new characteristic based on muhifractal spectrum is developed to describe the complexity of SEMG, it offers an alternate method for SEMG pattern classification.

关 键 词：多重分形 表面肌电信号 灰色关联 

分 类 号：R318[医药卫生—生物医学工程] TN911.7[医药卫生—基础医学]