一种新型小波阈值法在信号消噪中的应用研究

分析了sqtwolog、rigrsure、heursure、mini maxi4种常见的阈值确定方法后发现,信号中出现较高次谐波分量时,用sqtwolog和mini maxi阈值法去噪后高频信息损失严重,而用rigrsure和heursure法去噪后50Hz频率分量损失较为严重。为此,提出了一种新的阈值函数-先验知识的阈值确定方法,在信号去噪过程中运用先验知识阈值法确定各细节层的阈值来去噪。MATLAB仿真试验验证该方法有效性和优越性的结果表明:新的阈值法可以彻底的滤除信号中的噪声和无用的高频分量,尽可能的减少低频和有用高频信息的损失。

Application of an Improved Wavelet Threshold Method in De-noising Reduction

Based on the resolution analysis wavelet threshold de-noising methods which was put forward by D.L. Donoho and I. M. Johnstone, this paper introduces and compares four familiar getting threshold methods of the wavelet de-noising, i.e. sqtwolog, rigrsure, heursure and minimaxi. When the comparatively higher mode harmonic component appears in the signal, the high frequency information loses seriously by using the Sqtwolog and the Minimaxi threshold value laws, but the 50Hz component loses seriously by using the Heursure and the Rigrsure threshold value laws. This paper puts forward a new de-noising method--getting prior factors method. the new threshold value law can strain noise and useless high-frequency weight of signal with possible reduction of low frequency and loss of useful high-frequency information. The MATLAB simulation experiment indicates the new threshold value law may thorough filter in the signal the noise, and the numerical simulations show that this de-noising method is effective and excellent. Results also show that this method gets higher signal to noise ratio and better signal restoration when applying to the coherent noise corrupted digital signa1.