基于DSS和FSWT的欠定信号识别方法研究  

An efficient identification method for underdetermined signals based on DSS and FSWT

作　　者：王元生[1] 任兴民[1] 邓旺群[2] 杨永锋[1] 

机构地区：[1]西北工业大学振动工程研究所,西安710072 [2]中国航空动力机械研究所,湖南株洲412002

出　　处：《振动与冲击》2014年第21期80-84,共5页Journal of Vibration and Shock

基　　金：国家自然科学基金(11272257);陕西省自然科学基础研究(2011JQ1011);航空科学基金(20112108001);西北工业大学基础研究基金(JC201242)

摘　　要：针对旋转机械信号分析时产生的欠定信号盲源分离问题,建立了一种基于频率切片小波变换(Frequency Slice Wavelet Transformation,FSWT)和去噪源分离(Denoising Source Separation,DSS)的欠定信号分析方法(FSWT-DSS),首先通过FSWT反变换重构出新的混合信号,有效解决欠定盲分离维数不足的问题,再应用DSS分离得到源信号,解决了欠定盲分离问题,同时解决了单独应用FSWT时进行时频分析的不足。算法仿真和应用实例验证了FSWT-DSS方法在实测故障信号分析中的有效性。Combining the features of frequency slice wavelet transformation （FSWT ） and denoising source separation （DSS）,an efficient recognition method for underdetermined signals based on DSS and FSWT was proposed. This method was used to deal with the blind source separation （BSS）problem of rotating machineries in the case of the number of observed mixtures being less than that of contributing sources.New mixed signals could be reconstructed with the FSWT method,it was an effective way to solve the problem of the dimension insufficient in underdetermined blind source separation.At the same time,the method could be used to solve the problem of time-frequency analysis.Applying FSWT-DSS method to a rotor fault detection,the sudden unbalance phenomenon was diagnosed with the measured fault signals of the rotor.The simulation and experimental results showed that the FSWT-DSS method is indeed efficient in fault diagnosis and it has an important engineering significance for fault detection of rotating machines.

关 键 词：频率切片小波变换 欠定盲源分离 诊断 信号处理 去噪源分离 

分 类 号：TN911[电子电信—通信与信息系统]