欠定盲分离中源信号数目估计问题的研究

项目名称： 欠定盲分离中源信号数目估计问题的研究

项目编号： No.61203117

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 谭北海

作者单位： 广东工业大学

项目金额： 25万元

中文摘要： 盲信号分离因其巨大的潜在应用价值已成为国际上信号处理领域的热点研究问题，而欠定盲分离因其接近实际应用环境更是盲分离领域研究的焦点。以往欠定盲分离算法大多假设源信号数目已知条件下提出，因此如何估计信号源个数一直是一个十分棘手的问题。本项目将以模糊聚类分析、超平面聚类分析等手段对线性瞬时混叠欠定盲分离源信号数目的估计进行研究，且将研究延伸至卷积欠定盲分离源数估计，进而对此问题进行全面深入的探讨和分析，最后还从微弱信号处理如脑电信号处理、胎儿心电信号处理等去验证本项目成果的有效性和可靠性，为欠定盲分离的研究和应用奠定基础。

中文关键词： 欠定盲分离；源数估计；稀疏表示；胎心电；

英文摘要： Because of its great potential value of application, blind signal separation has become a hot topic in the field of signal processing. Meanwhile underdetermined blind source separation, with its being in conformity with actual environment, has been the focus of research in the field of blind source separation. In the past, the algorithms of underdetermined blind source separation were proposed when they suppose the number of source signals was known, so how to estimate the number of source is a very difficult problem. This project will make researches on estimation of the number of source signals of the linear instantaneous mixed underdetermined blind source separation by means of fuzzy clustering analysis,hyperplane clustering analysis etc.. At the same time,estimation of the number of source signals of convoluted underdetermined blind separation are also studied in this project.Furthermore, more comprehensive and in-depth discussion and analysis will be made. Finally, the validity and reliability of the achievement will be verified by weak signals processing,such as EEG processing, FECG processing. All the results in this project will build foundation for researches and applications of underdetermined blind source separation in the future.

英文关键词： underdetermined blind source separation；source estimation；sparse representation；FECG；