研究零相位数字滤波器在进行非平稳信号滤波时的特点.选用一种典型带通零相位数字滤波器,以非平稳调幅信号作为滤波器输入,进行仿真分析.将零相位数字滤波器与小波包分解重构和经验模态分解方法的滤波能力进行了比较.探讨了零相位数字滤波器在处理非平稳调幅信号时存在的过渡过程,及对滤波结果幅值和频率的影响.论述了滤波误差与滤波器参数、输入信号特性和信噪比等因素的关系.提出了分段零相位滤波器方法,改善了滤波器性能.最后,以一个实测的振荡信号对上述分析进行了验证.为在非平稳信号处理中,正确使用零相位数字滤波器提供了参考.
The characteristics of zero-phase digital filter when being utilized to process the non-stationary signals are studied. Typical band-paess zero-phase digital filters are simulated. And non-stationary amplitude-modulation signals are selected to be input of the simulation. Wavelet packet decomposition and reconstruction, empirical mode decomposition and the zero-phase digital filter are applied to com- pare their band-pass filter capabilities. The simulation demonstrates the transition process in non-sta- tionary signal filtering. And it clarifies amplitude characteristics, and frequency characteristics existing in filtering the non-stationary amplitude-modulation signal in detail. This article also discusses the rela- tionship between error and filter parameters, characteristics of input signal and signal to noise ratio of input signal. A segment zero-phase digital filter method is proposed in this paper to enhance the perfor- mance of the normal zero-phase digital filter. The segment zero-phase digital filter is employed in ex- tracting the main component from a real oscillation signal to verify the validity of the new zero-phase filtering method. The study provides support for proper usage of zero-phase digital filter applied on non-stationary signal processing.