摘要
本文提出采用双维谱分析方法对非高斯分布的随机信号进行研究,并对双维谱分析原理作了较系统的论述,采用计算机进行了分析和运算。由计算机仿真的结果,表明了双维谱分析方法能有效地检测出仅由频率成分间的相对相位变化表现的机械系统的微细不正常状态,并且能够有效地抑制有害的附加的高斯噪声。
In this paper, we propose a bispectral analysis method for studying the non-Gauss distributed random signals. The principle of this method is discussed systematically and the program for computation of bispectra is finished by using a microcomputer. The results of computer simulation analyses show that the bispectral analysis method is effective to detect even such a delicate abnormality of machine system that appears only as the change of the relative phase among frequency components and it can also suppress the corrupty additive Gaussian noise.
出处
《同济大学学报(自然科学版)》
EI
CAS
CSCD
1990年第3期347-354,共8页
Journal of Tongji University:Natural Science
基金
国家自然科学基金
关键词
双维谱
设备
噪声
诊断
随机信号
Random signal analysis
Fault diagnosis
Bispectrum