从噪声中提取瞬态信号──自适应权向量法  

Adaptive-Weight-Vector Method for Detecting Transient Signal

作　　者：董大群[1] 曾南[1] 李斌[1] 

机构地区：[1]西北工业大学

出　　处：《西北工业大学学报》1994年第3期347-353,共7页Journal of Northwestern Polytechnical University

摘　　要：本文提出一种自适应方法，从宽带噪声中检测瞬态信号．由于检测是在自适应权向量空间进行的，因此称之为自适应权向量法（ＡＷＶＭ）．文中对ＡＷＶＭ的机理进行了分析．仿真结果证实了理论分析的正确性。In the authors engineering experience, detection of transient signal can be done with methods that arc nominally unrestricted but really only applicable to the case in which the signal is stationary. Porat et al in 1986 made a good start on how to detect transient signal [4]. In this paper the authors propose a method which is different from that of Ref. [4] and is believed to be capable of such detection in engincering practice.Transient signal is characterized by its conspicuous nonstationarity. Obviously, the detection cannot wait until the nonstationary process is ended; thus the detection must utilize something in the transient leaning process. That something, in the authors proposed method, is the excess weight vector which is a dynamic image of the transfen t signal. The proposed method is called by the authors the AWVM(Adaptive -Weight-Vector Method). Eq. (21) gives the output for step input of the AWVM detection system. Simulation results indicate that the theoretical analysis leading to the AWVM is trustworthy and that the AWVM detection system can apprccially raise the signal -to-noise ratio.

关 键 词：自适应权向量 瞬太信号 噪声 

分 类 号：TN911.23[电子电信—通信与信息系统]