摘要
针对传统傅里叶谱估计方法不适合非平稳信号和对短信号谱分析能力弱的问题,提出一种基于边际谱的功率谱估计方法,给出了边际谱的求解过程,证明了边际谱的线性性质,并在与Fourier谱比较的基础上给出了边际谱的物理意义和频率分辨率。详细讨论了谱估计的算法和步骤,并用短信号和非平稳信号进行了仿真验证。理论与仿真结果表明:在短信号和非稳态信号情况下,基于边际谱的功率谱估计方法对数据要求更短,且其频率分辨率和估算精度都要高于傅里叶方法。
FFT method can not meet the basic requirements of power spectrum for non-stationary signal and short signal.A new spectrum estimation method based on marginal spectrum from Hilbert-Huang transform（HHT） was proposed.The procession of obtaining marginal spectrum in HHT method was given and the linear property of marginal spectrum was demonstrated.Compared with the FFT method,the physical meaning and the frequency resolution of marginal spectrum were further analyzed.Then the Hilbert spectrum estimation algorithm was discussed in detail,and the simulation results were given at last.The theory and simulation shows that under the condition of short data signal and non-stationary signal,the frequency resolution and estimation precision of HHT method is better than that of FFT method.
出处
《核电子学与探测技术》
CAS
CSCD
北大核心
2011年第9期1062-1066,共5页
Nuclear Electronics & Detection Technology
基金
国家863计划(2006AA06Z105)
湖南省"十一五"重点建设学科-光学基金
湖南省教育厅项目(08C615)资助
