基于FIR数字滤波器随机振动信号频时域转换方法

Transformation from Frequency to Time Domain of Random Vibration Signal Based on FIR Digit Filter

基于有限长单位冲激响应数字滤波器,根据指定ESA功率谱密度函数特征,设计了FIR数字滤波器,并通过控制其参数,获得了满足预定条件和要求的滤波器幅相特性.利用白噪声序列作为滤波器输入端,生成了符合要求的随机振动时域信号.同时还采用基于AR自回归模型参数方法对获得的随机振动时域信号进行功率谱估计与验证.结果表明:功率谱精度及两功率谱曲线间的误差完全在允许范围内,采用该方法获得的随机振动时域信号完全满足仿真与试验测试对信号幅频特性和功率的要求.

In terms of the given ESA power spectral density（PSD） function characteristics,a FIR digit filter was designed based on finite unit impulse response.The filter amplitude-phase characteristics that completely met the conditions of given power spectral characteristics were obtained by changing parameters of FIR digit filter.The random vibration time signal was generated with the white noise as input of the filter.Furthermore,autoregressive parameter model was used to estimate and validate the power spectrum of obtained time signal.The results show that the error between two spectral curves are in the range of permits,and the random vibration time signal obtained using presented method could completely meet the requirements of simulation and test.