Blind Sources Separation (BSS) provides a new alterative for extraction of certain signal components in the signal corrupted by noise, and presents a new method for mechanical fault diagnosis. The basic theories about BSS are reviewed and a BSS algorithm for detection and control criteria based on the VQM threshold is proposed. The effectiveness of the BSS algorithm is verified through extraction of the transients in the signal produced by convolution of transients and noise.
【更新日期】
2006-05-31
【分类号】
TB533
【正文快照】
0引言在复杂环境背景下,传感器获得的信号通常是多个源信号的混合信号。在不知道任何预知的背景知识下,仅根据各信号源相互统计独立的假设,从多个传感器的测量信号中恢复出各个源信号的技术便是盲分离(Blind Sourse Separation或Blind Signal Separation,简称BSS)技术。BSS已经