一种有效实现弱平稳随机信号完整功率谱估计的方法

A effective method of estimating the weak-stationary random signal power

为解决某些振动信号中含有微弱频率难分辨的问题,本文采用现代信号处理的小波分析和MUSIC(多重信号分类)谱估计,提出了一种新的实现弱平稳随机信号完整功率谱估计的方法。该方法将弱平稳随机信号利用小波分解与重构得到各频段的细化信号,再将经MUSIC谱估计后的细化功率谱融合,得到具有较高分辨率的完整弱平稳随机信号功率谱。

Some weak vibration signals arc very difficult to iolentify in frequency domain , the paper introduces a new signal processing method for estimating the weak-stationary random signal power with wavelet and MUSIC（Multiple Signal Classification） .First to use the wavelet to dissolve and recompose weak-stationary random signal , then use the MUSIC to estimate power density of the every frequency domain signal from the wavelet analysis. Finally add together all domain power density to get the entire weakstationary random signal power spectrum with better resolution.