基于小波变换的汽轮机振动信号软阀值消噪技术研究

A Study on Signal Thresholding De-noising Technique Based on the Wavelet Transform

信号消噪是小波变换的重要应用,介绍了小波消噪的基本原理及其主要步骤,以及软阀值消噪方法及软阀值规则的选取,最后利用实验仿真信号和现场实测汽轮机振动信号,并考虑噪声方差估计的对消噪的影响,分析比较了各种软阀值选取方式的消噪效果,从而得到软阀值选取方式的有价值的规律和原则,利用最佳软阀值处理后的小波系数重建信号,分析结果表明该方法能够最有效地消除噪声。

Signal de-noising is an important applications on wavelet transform technology. The basis theory and process of wavelet de-noising are introduced firstly in this paper, the de-nosing method of the soft thresholding and the choices of soft thresholding rule are analyzed particularly. Considering the de-noising influence of noise variance estimation. Some valuable rules of the choice of soft thresholding was discovered by the analysis of simulated signal and local vibration fault signal of the turbine, original signal is reconstructed utilizing wavelet coefficient of the best treated with thresholding, which prove that the method is the most effective for de-noising.