面向卷积混叠环境下的盲源分离新方法  

Novel Blind Source Separation Method for Convolutive Mixed Environment

作　　者：解元 邹涛 孙为军 谢胜利 XIE Yuan;ZOU Tao;SUN Wei-Jun;XIE Sheng-Li(School of Mechanical and Electrical Engineering,Guangzhou University,Guangzhou 510006;Key Laboratory of Intelligent Information Processing and System Integration of Internet of Things,Ministry of Education,Guangzhou 510006;Guangdong Provincial Key Laboratory of Information Technology of Internet of Things,Guangzhou 510006;Key Laboratory of Intelligent Detection and the Internet of Things in Manufacturing,Ministry of Education,Guangzhou 510006;Discrete Manufacturing Intelligence Discipline Innovation and Talent Introduction Base Based on Internet of Things Technology,Guangzhou 510006;Guangdong-Hong Kong-Macao Joint Laboratory for Smart Discrete Manufacturing,Guangzhou 510006)

机构地区：[1]广州大学机械与电气工程学院,广州510006 [2]物联网智能信息处理与系统集成教育部重点实验室,广州510006 [3]广东省物联网信息技术重点实验室,广州510006 [4]智能检测与制造物联教育部重点实验室,广州510006 [5]基于物联网技术的离散制造智能化学科创新引智基地,广州510006 [6]粤港澳离散制造智能化联合实验室,广州510006

出　　处：《自动化学报》2023年第5期1062-1072,共11页Acta Automatica Sinica

基　　金：国家重点研发计划(2018YFB1802400);国家自然科学基金(62003095,52171331)资助。

摘　　要：卷积混叠环境下的盲源分离(Blind source separation, BSS)是一个极具挑战性和实际意义的问题.本文在独立分量分析框架下,建立非负矩阵分解(Nonnegative matrix factorization, NMF)模型,设计新的优化目标函数,通过严格的数学理论推导,得到新的模型参数更新规则;并对解混叠矩阵进行标准化处理,避免幅度歧义性问题;在源信号的重构阶段,通过实时更新非负矩阵分解模型参数,避免源信号的排序歧义性问题.实验结果验证了所提算法在分离中英文语音混叠信号、音乐混叠信号时的有效性和优越性.Blind source separation(BSS)for convolutive mixed environment is a challenging and practical topic.In this paper,a nonnegative matrix factorization(NMF)model is established based on the framework of independent component analysis,and a new optimization objective function is designed.Through strict mathematical theory derivation,new model parameters update rules are obtained,and the demixing matrix is standardized to avoid the scale ambiguity.In the stage of source reconstruction,the permutation ambiguity can be avoided by updating the parameters of the NMF model in real time.Experimental results verify the effectiveness and superiority of the proposed algorithm in separating Chinese speech mixtures,English speech mixtures,and music signal mixtures.

关 键 词：盲源分离 卷积混叠 独立分量分析 非负矩阵分解 

分 类 号：TN911.7[电子电信—通信与信息系统]