基于Matlab自然伽马测井曲线小波降噪算法研究

通过对小波变换算法的理论分析,建立了小波降噪算法的判定准则和信号降噪模型。通过对信号降噪常用方法分析,采用快速傅里叶变换FFT和小波阈值降噪方法对自然伽马测井曲线进行降噪。在小波阈值降噪方法中,采用3种从原始信号确定阈值(全局阈值、penalty阈值和Birge-Massart阈值)和4种基于样本估计的阈值(rigrsure阈值、sqtwolog阈值、heursure阈值和minimaxi阈值),对降噪效果进行分析,小波阈值降噪方法明显优于快速傅里叶变换FFT降噪方法。通过信噪比、相关系数、降噪信号能量比和降噪信号与原信号均方差4个性能指标对比,采用rigrsure自适应软阈值小波降噪方法对自然伽马测井曲线降噪,能有效去除自然伽马测井曲线中与地层信息无关的统计涨落和毛刺干扰,得到最佳的降噪效果。该方法在自然伽马测井曲线降噪处理中得到很好地应用。

On Natural Gamma Logging Curve Wavelet De-noising Algorithm Based on Matlab

The theory of wavelet transform algorithm is analyzed,and the judgment criterion of wavelet de-noising algorithm and the signal de-noising model are established.At the same time,the wavelet threshold noise reduction is studied and discussed.By analyzing the common methods of signal de-noising,methods of Fast Fourier Transform and wavelet threshold de-noising are used to de-noise and compared.In the wavelet threshold de-noising method,three kinds of determined threshold value from the original signal (global threshold,penalty threshold and Birge-Massart threshold) and four kinds of sample-based estimation of the threshold (rigrsure threshold,sqtwolog threshold,heursure threshold and minimaxi threshold) are used to compare and analyze the de-noising effects.Wavelet threshold de-noising method is superior to the Fast Fourier Transform de-noising method.Four performance indicators of the signal-to-noise ratio,correlation coefficient and noise signal energy ratio and the original signal mean square error are compared.Rigrsure adaptive soft threshold wavelet de-noising method for natural gamma ray logging curve is used.This method can effectively remove the statistical fluctuation and the burr interference in the natural gamma logging curve,it can obtain best noise reduction effect and get good application in the natural gamma logging curve noise reduction processing.