本文主要研究内容
作者张汉勇(2019)在《基于加权相似性网络和水平可视图的脑电信号分析方法研究》一文中研究指出:生物信号是现代数字信号处理的重要研究领域。由于生命体系统的复杂性,生物信号通常具有非线性、混沌、非平稳等特点,属于非线性时间序列。正是因为这些特点,使得生物信号分析区别于一般的信号处理,成为现代数字信号处理的一个重要分支。时间序列到复杂网络转换算法的提出,使得复杂网络成为时间序列分析的新工具。复杂网络经过多年的发展,形成了成熟的理论体系,在诸多学科领域得到了大规模应用。本文在已有的复杂网络构造方法的基础上,主要对加权相似性网络和加权水平可视图进行深入研究。本文总结了已有的相似性网络和加权水平可视图构造方法,通过对当前算法的不足之处进行改进,提出了新的复杂网络构造方法,即改进的加权相似性网络和改进的加权水平可视图方法。论文主要方法和创新点如下:(1)提出了改进的加权相似性网络算法。传统的相似性网络在构造过程中由于节点数量较少,同时没有引入边权值信息,会造成时间序列信息的丢失。把癫痫脑电时间序列平均分段之后,对每个子段时间序列提取12个时域统计特征作为复杂网络的节点,使得节点数量增加;计算节点间的欧氏距离,设定阈值,确定节点间的连边关系,并把节点间的斜率作为连边权值,构造加权邻接矩阵,完成癫痫脑电信号加权相似性网络的构造。提取平均权重度作为分类特征,用于癫痫发作期信号和癫痫发作间歇期信号的分类,分类准确率最高为96%。实验结果表明改进的加权相似性网络的性能较之传统的相似性网络算法有了很大提升。(2)传统的加权水平可视图算法不能很好反映时间序列的动力学特性,因此,本文提出了三种加权水平可视图改进算法:一是使用12个时域统计特征作为加权水平可视图的节点,以节点间的差值作为连边权值,构造加权水平可视图,降低了算法的时间复杂度;二是使用了新的连边权值,节点间的斜率和节点间的角度作为连边权值,构造加权水平可视图,斜率和角度更能反映时间序列隐藏的动力学特性;三是使用原癫痫脑电时间序列作为节点,以节点间的斜率作为正向网络连边权值,以节点间的角度作为反向网络连边权值,首次提出了双向加权水平可视图算法。随后将改进的加权水平可视图算法用于癫痫脑电信号分类。在癫痫脑电加权水平可视图构造完成之后,提取平均权重度、权重分布熵等分类特征,实现癫痫发作期信号和癫痫发作间歇期信号的准确分类,分类最高准确率为98.5%。实验结果表明改进的加权水平的性能较之传统的复杂网络算法有了很大提升。(3)本文提出了改进的加权相似性网络和加权水平可视图构造算法,并应用这两个复杂网络构造算法将癫痫脑电信号转换为对应的复杂网络,提取复杂网络拓扑统计特征作为分类特征,实现了高质量的癫痫脑电信号分类,具有重要的临床研究意义。通过和其他相关实验的比较,验证了本文所提出的算法可以获取生物信号更全面、更深层的动力学特征,说明本文算法可以有效应用于生物信号分析。
Abstract
sheng wu xin hao shi xian dai shu zi xin hao chu li de chong yao yan jiu ling yu 。you yu sheng ming ti ji tong de fu za xing ,sheng wu xin hao tong chang ju you fei xian xing 、hun dun 、fei ping wen deng te dian ,shu yu fei xian xing shi jian xu lie 。zheng shi yin wei zhe xie te dian ,shi de sheng wu xin hao fen xi ou bie yu yi ban de xin hao chu li ,cheng wei xian dai shu zi xin hao chu li de yi ge chong yao fen zhi 。shi jian xu lie dao fu za wang lao zhuai huan suan fa de di chu ,shi de fu za wang lao cheng wei shi jian xu lie fen xi de xin gong ju 。fu za wang lao jing guo duo nian de fa zhan ,xing cheng le cheng shou de li lun ti ji ,zai zhu duo xue ke ling yu de dao le da gui mo ying yong 。ben wen zai yi you de fu za wang lao gou zao fang fa de ji chu shang ,zhu yao dui jia quan xiang shi xing wang lao he jia quan shui ping ke shi tu jin hang shen ru yan jiu 。ben wen zong jie le yi you de xiang shi xing wang lao he jia quan shui ping ke shi tu gou zao fang fa ,tong guo dui dang qian suan fa de bu zu zhi chu jin hang gai jin ,di chu le xin de fu za wang lao gou zao fang fa ,ji gai jin de jia quan xiang shi xing wang lao he gai jin de jia quan shui ping ke shi tu fang fa 。lun wen zhu yao fang fa he chuang xin dian ru xia :(1)di chu le gai jin de jia quan xiang shi xing wang lao suan fa 。chuan tong de xiang shi xing wang lao zai gou zao guo cheng zhong you yu jie dian shu liang jiao shao ,tong shi mei you yin ru bian quan zhi xin xi ,hui zao cheng shi jian xu lie xin xi de diu shi 。ba dian xian nao dian shi jian xu lie ping jun fen duan zhi hou ,dui mei ge zi duan shi jian xu lie di qu 12ge shi yu tong ji te zheng zuo wei fu za wang lao de jie dian ,shi de jie dian shu liang zeng jia ;ji suan jie dian jian de ou shi ju li ,she ding yu zhi ,que ding jie dian jian de lian bian guan ji ,bing ba jie dian jian de xie lv zuo wei lian bian quan zhi ,gou zao jia quan lin jie ju zhen ,wan cheng dian xian nao dian xin hao jia quan xiang shi xing wang lao de gou zao 。di qu ping jun quan chong du zuo wei fen lei te zheng ,yong yu dian xian fa zuo ji xin hao he dian xian fa zuo jian xie ji xin hao de fen lei ,fen lei zhun que lv zui gao wei 96%。shi yan jie guo biao ming gai jin de jia quan xiang shi xing wang lao de xing neng jiao zhi chuan tong de xiang shi xing wang lao suan fa you le hen da di sheng 。(2)chuan tong de jia quan shui ping ke shi tu suan fa bu neng hen hao fan ying shi jian xu lie de dong li xue te xing ,yin ci ,ben wen di chu le san chong jia quan shui ping ke shi tu gai jin suan fa :yi shi shi yong 12ge shi yu tong ji te zheng zuo wei jia quan shui ping ke shi tu de jie dian ,yi jie dian jian de cha zhi zuo wei lian bian quan zhi ,gou zao jia quan shui ping ke shi tu ,jiang di le suan fa de shi jian fu za du ;er shi shi yong le xin de lian bian quan zhi ,jie dian jian de xie lv he jie dian jian de jiao du zuo wei lian bian quan zhi ,gou zao jia quan shui ping ke shi tu ,xie lv he jiao du geng neng fan ying shi jian xu lie yin cang de dong li xue te xing ;san shi shi yong yuan dian xian nao dian shi jian xu lie zuo wei jie dian ,yi jie dian jian de xie lv zuo wei zheng xiang wang lao lian bian quan zhi ,yi jie dian jian de jiao du zuo wei fan xiang wang lao lian bian quan zhi ,shou ci di chu le shuang xiang jia quan shui ping ke shi tu suan fa 。sui hou jiang gai jin de jia quan shui ping ke shi tu suan fa yong yu dian xian nao dian xin hao fen lei 。zai dian xian nao dian jia quan shui ping ke shi tu gou zao wan cheng zhi hou ,di qu ping jun quan chong du 、quan chong fen bu shang deng fen lei te zheng ,shi xian dian xian fa zuo ji xin hao he dian xian fa zuo jian xie ji xin hao de zhun que fen lei ,fen lei zui gao zhun que lv wei 98.5%。shi yan jie guo biao ming gai jin de jia quan shui ping de xing neng jiao zhi chuan tong de fu za wang lao suan fa you le hen da di sheng 。(3)ben wen di chu le gai jin de jia quan xiang shi xing wang lao he jia quan shui ping ke shi tu gou zao suan fa ,bing ying yong zhe liang ge fu za wang lao gou zao suan fa jiang dian xian nao dian xin hao zhuai huan wei dui ying de fu za wang lao ,di qu fu za wang lao ta pu tong ji te zheng zuo wei fen lei te zheng ,shi xian le gao zhi liang de dian xian nao dian xin hao fen lei ,ju you chong yao de lin chuang yan jiu yi yi 。tong guo he ji ta xiang guan shi yan de bi jiao ,yan zheng le ben wen suo di chu de suan fa ke yi huo qu sheng wu xin hao geng quan mian 、geng shen ceng de dong li xue te zheng ,shui ming ben wen suan fa ke yi you xiao ying yong yu sheng wu xin hao fen xi 。
论文参考文献
论文详细介绍
论文作者分别是来自济南大学的张汉勇,发表于刊物济南大学2019-10-31论文,是一篇关于相似性网络论文,加权水平可视图论文,生物信号论文,癫痫脑电论文,自动检测分类论文,济南大学2019-10-31论文的文章。本文可供学术参考使用,各位学者可以免费参考阅读下载,文章观点不代表本站观点,资料来自济南大学2019-10-31论文网站,若本站收录的文献无意侵犯了您的著作版权,请联系我们删除。
标签:相似性网络论文; 加权水平可视图论文; 生物信号论文; 癫痫脑电论文; 自动检测分类论文; 济南大学2019-10-31论文;