| Wavelet analysis developed in recent years is a new mathematic branch and is being widely used in many scientific fields. This method is characterized by fine localized time and frequency domain that could converge signals at any detail and its time domain resolution could change and increase with scale factor. It provides many advantages to application in well logging signals processing. This paper does some in depth researches in its applications and gets three important outcomes:1. Improving vertical resolution of AC well logging curves. Use the method of Mallat arithmetic to decompose well logging curves. Signals can be decomposed into two parts, approximation and detail. After three-point deconvolution on the detail part, the reconstructed curves get higher resolution, and the influence of surrounding rock decrease, true logging value can be got, and then we can calculate more accurate porosity value.2. Present a new method for detecting the signal start points in full wavetrain acoustic component wave. Adapting Gauss complex wavelet, we studied the characters of equation of these signals' continuous wavelet transform, then construct a detecting function, which is the ratio of module of wavelet coefficient between two different scale. This function decreases the affection of amplitude and strengthens the characters of frequency. Thus, wave fronts of the sliding P-waves, the sliding S-waves and the Stoneley-waves can be estimated precisely.3. Use Morlet wavelet to analysis the demarcation of cycle of sedimentation. By continuous wavelet transform of GR curve, we can find time-frequency characters in complicate logging signals. After some experiments, a way of selecting the best scale is found. And in the best scale, we could analyze the cycle of well logging curve, and then demarcate the cycle of sedimentation more precisely and visually with GR, RT curves and so on.
|
PDF Full Text Request