基于运动想象的脑机接口的数学模型与算法研究

发布时间：2018-02-13 07:09

本文关键词： BCI 支持向量机一阶μ波二阶μ波离散度 CSP分类法　出处：《长沙理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：脑-机接口(Brain-computer interface,BCI)是在大脑和外部设备之间建立的一种信息通道,传递人的需求信息.它可以把人的需求直接转换成外部指令,从而代替人的肢体或语言并实现与外界的交流.BCI这一技术将为更多的瘫痪、聋哑患者提供便捷的沟通工具.本文主要研究基于运动想象的BCI数学模型与算法研究,主要目的是寻找和定义最能区分不同类型运动想象脑电信号的数量特征,并构建恰当的分类算法,得到较高的识别分辨率.全文共分为七章.第一章主要介绍BCI的背景及前景,并介绍了本文的主要工作.第二章介绍与本文研究相关的基础知识,包括EEG的频域分析法等.第三章介绍基于一阶μ波水平的电极筛选和特征分类,然后进行任务相关分析,进而根据第一平均值上下界定法和第二平均值上下界定法的出相应的准确率.第四章主要讨论二阶μ波水平的概念,以及用二阶μ波分析实验数据.第五章主要介绍l1-距离的概念及算法实现.第六章主要介绍l2-距离的概念及算法实现.第七章分为三个部分,包括CSP分类算法的简介,实验数据分析及结果分析及总结
[Abstract]:Brain-Computer Interface (BCI) is a channel of information between the brain and external devices that conveys human needs. It converts human needs directly into external instructions. In order to replace the human body or language and achieve communication with the outside world, this technology will provide more paralyzed, deaf and mute patients with convenient communication tools. This paper mainly studies the BCI mathematical model and algorithm based on motion imagination. The main purpose of this paper is to find and define the best quantitative features of different types of motion imaginary EEG signals, and construct an appropriate classification algorithm to obtain higher recognition resolution. The whole paper is divided into seven chapters. Chapter one mainly introduces the background and prospect of BCI. The main work of this paper is introduced. Chapter two introduces the basic knowledge related to this paper, including the frequency domain analysis of EEG, etc. Chapter 3 introduces the electrode selection and feature classification based on the first order 渭 wave level, and then carries on the task correlation analysis, the third chapter introduces the electrode selection and the characteristic classification based on the first order 渭 wave level, then carries on the task correlation analysis. Then according to the corresponding accuracy of the first average upper and lower definition method and the second average upper and lower definition method. Chapter 4th mainly discusses the concept of second order 渭 wave level. Chapter 5th mainly introduces the concept of l1-distance and its algorithm implementation. Chapter 6th mainly introduces the concept of L2-distance and its algorithm implementation. Chapter 7th is divided into three parts, including the introduction of CSP classification algorithm. Analysis and summary of experimental data and results
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R338;TP18;TN911.7

【参考文献】