用于动态测试的信号分析方法  

SIGNAL PROCESSING METHODS IN DYNAMIC MEASUREMENT

作　　者：张朝晖[1] 袁建美[2] 蒋洪明[3] 

机构地区：[1]石油大学信息与控制工程学院,山东东营257061 [2]石油大学应用数学系,山东东营257061 [3]东南大学仪器系,江苏南京210096

出　　处：《石油大学学报（自然科学版）》2002年第2期120-124,共5页Journal of the University of Petroleum,China(Edition of Natural Science)

基　　金：石油大学基础研究基金 (ZX9914 );博士科研基金资助

摘　　要：信号分析被广泛用于消噪、特征提取、状态识别、故障诊断等。常用的方法有参数化分析法和非参数化分析法。参数化分析方法有ARMA模型、分形维数计算及用于判别微弱信号存在性的混沌分析法等 ;非参数化方法即谱分析方法 ,包括传统的DFT谱方法及在此基础上的提高频率分辨能力 (如AR功率谱 )和时频联合分析 (小波分解、Wigner谱 )等内容。文中对各种方法的原理、特点作了综述 。Signal processing algorithms have been broadly used for cancelling noise, extracting feature, identifying status and diagnosing malfunction. Many branches of signal processing methods have been developed, including parametric analysis and non parametric analysis. The parametric analyses include ARMA model, fractal dimension, and chaotic analysis for detecting weak signals in strong noise environments. The non parametric analyses include regular DFT spectrum, high resolution spectrum (AR power spectrum), frequency/time joint analyses (wavelets,Wigner distributions). The principles of those methods are reviewed, and their characters are compared with each other. Their applications are exemplified.

关 键 词：动态测试 频域分析 时域分析 时-频域分析 分形维数 混沌分析 参数化分析法 非参数化分析法 动态测试 信号分析方法 

分 类 号：TN911.6[电子电信—通信与信息系统]