摘要
对于非平稳随机信号的功率谱分析 ,采用自相关函数傅里叶变换和直接傅里叶变换分析时 ,由于存在数据截断 ,二者从概念上都不能很好地反映随机信号的功率谱。采用参数谱估计方法 ,则不能实现等百分比带宽分析。对于噪声和结构振动等信号 ,需要进行等百分比带宽分析。采用滤波法进行功率谱分析 ,不存在数据截断带来的误差 ,还可方便地用于等百分比带宽的谱分析。利用数字滤波器可以实现时分和频分复用的特点 ,结合重抽样技术来实现整个频率轴上的实时功率谱分析。上述算法可以通过高速数字信号处理器 (DSP)芯片实时地完成。文中介绍了基于数字滤波的低频随机信号实时功率谱分析仪的实现方法 ,并给出了实时算法的设计要点。
The correlation method and FFT method of the power spectrum estimation could not completely reflect the concept of the power spectrum of the non-stationary random signals because of data truncation. And the method of parametric spectrum estimation could not realize the analysis of a fixed percentage bandwidth which is necessary for noise signal, the structure vibration signal, etc. But the filter method avoids the error caused by truncated data and it can achieve the analysis of a fixed percentage bandwidth easily. The digital filter can realize the time division multiplexing and the frequency division multiplexing,so the real time power spectrum estimation can be achieved completely on the whole frequency axis with integrated resampling technique. This algorithm has been realized on high speed DSP chip in real time. In this paper, a real time power spectrum analyzer of low frequency random signal based on digital filters is introduced, and a design method of real time arithmetic is given.
出处
《电子工程师》
2005年第2期1-3,16,共4页
Electronic Engineer
关键词
实时功率谱分析仪
DSP
数字滤波
重抽样技术
real time power spectrum analyzer, DSP, digital filter, resampling technique
