Application of Numerical Methods Based on Wavelet Transforms for Detection of Homogeneous Areas on Logging Diagrams

The existing algorithms for filtering and selecting homogeneous intervals of logging diagrams in both depth and time scales are described. A system of criteria is developed, which allows to make a mathematical formulation of the problem and compare different approaches to log data processing. The method of applying the numerical scheme of selecting homogeneous intervals in the logging material on the basis of wavelet transformations is presented in detail. The software implementation of the proposed approach was carried out and a number of numerical experiments on the use of wavelet transforms for solving the set problems on the example of high-frequency induction logging isoparametric sounding signals were performed. In particular, the algorithm was proposed, which, according to experts-geophysicists, shows better performance on complex data and has more flexible tuning capabilities to take into account a priori information. A good result was obtained for automatic identification of logging equipment stop intervals for the primary recorded signals in the time scale.