盲源分离算法在混合震动信号分离中的应用  

The application of blind source separation in mixed vibration signal separation

作　　者：张林颖[1] 苏新彦[1] 韩焱[1] 姚金杰[1] 

机构地区：[1]中北大学信息与通信工程学院,山西太原030051

出　　处：《电子技术应用》2014年第7期137-140,共4页Application of Electronic Technique

基　　金：国家973计划前期研究专项(2011CB311804)

摘　　要：提出了一种基于z变换域有理传递函数F的时间延迟正定盲源分离算法,并提出将其应用于人工勘探地震波时传感器采集到的混合震动信号的信噪分离及横、纵波的分离。该算法既适用于分离以线性方式混合的信号,也适用于以非线性方式混合信号。时间延迟长短的选择依赖于有待处理的震动信号数据的长度。仿真结果表明,该算法能有效地对Matlab生成的人工模拟震动波进行信噪分离及横、纵波的分离,为震动信号数据的后期处理及分析提供有利依据。This paper proposes a time delay definite blind source separation algorithm based on rational transfer function in z- transform domain and uses this algorithm to separate signal and noise, transverse wave and longitudinal wave of mixed signal collected by sensors when we explorate the seismic wave. This algorithm is suitable for separating mixed signals in linear nonlinear. Selecting the length of the time delay depends on the length of the vibration signal data to be processed. Simulation results show that the algorithm can effectively separate signal and noise, transverse wave and longitudinal wave of artificial simulation shock wave generated by Matlab, which provides a favorable basis for post-processing and analysis of vibration signal data.

关 键 词：线性 非线性 盲源分离 人工地震勘探 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]