摘要
非平稳信号的分析越来越受到人们的重视。短时傅里叶变换(STFT)是一种线性变换,避免了其他高次型非平稳信号分析方法中出现的交叉项的干扰,是分析非平稳信号的有力工具。短时傅里叶变换的基本思想是利用一个固定大小的滑动的窗口函数对信号进行分析,并假定信号在窗口内是平稳的,因此加窗变换所固有的混叠现象在短时傅里叶变换中依然存在。本文基于频域插值的思想,提出了基于频域内插抗混叠短时傅里叶变换的算法,首先分析了混叠产生的物理本质,然后以汉明窗为基础构造了频域内插的方法,并利用牛顿插值法得出插值的迭代求解方法,最后给出内插优化算法步骤。仿真实例验证了算法的可行性。
More and more researchers attached much importance to the analysis of the nonstationary signals, Because Short Time Fourier Transform （STFT） is a kind of linear transform, which avoids the disturbances of crossover terms, STFT becomes one of the most powerful tools in analysis of the nonstatlonary signals. The kernel idea of STFT is to use a size-fixed and moving ＂window＂ to analyze the nonstationary signals, and at the same time, the idea supposes that the signal in the ＂window＂ is stationary. Because of the inherent shortcoming, aliasing still exists in the process of STFT. Based on the interpolation in the frequency domain, a kind of anti-aliasing algorithm using STFT in the frequency domain is brought forward in this paper. At first the physical essence of STFT aliasing is analyzed, then a frequency domain interpolation method is proposed by choosing Hamming window, and using Newton interpolation theory the iterative steps of interpolation is given. Finally the implementation steps of the algorithm are listed. Simulation results show that the proposed algorithm can effectively eliminate the STFT aliasing in nonstationary signal detection.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2008年第2期284-288,共5页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(90407007,60372001)
教育部博士点专项基金(20030614006)资助项目
关键词
短时傅里叶变换
内插
非平稳信号检测
抗混叠
迭代算法
short time Fourier transform （STFF）
interpolation
nonstationary signal detection
anti-aliasing
iterative algorithm
