Picking Up Information From Well Log Data

Guided by the theory of sequence stratigraphy, new sedimentary indexes are found in the well-logging data on the base of sedimentary theory, wavelet analysis, fractal theory and Principal Component Analysis, etc. And the geological model is set up based on the new indexes. Then the recognition of storing beds and sedimentary phases is measured and automatic. Also the previous well-logging data processing, the method of picking up the sedimentary characteristics and the measured assessment model of the storing beds are established in this thesis. The principal component analysis can pick up few irrelative principal components that can reflect stratum characteristics from relative well logging parameters. And the principal components can reflect the stratum characteristics synthetically and effectively. Using the principal components can reach the goal of measuring the stratum effectively. In the well-logging data noises exist inevitably. In this thesis wavelet denoising is adopted and suitable thresholds are chosen to deal with the coefficients after the well-logging data is decomposed by wavelet. After wavelet denoising, the noise intention is reduced greatly. The fractal dimensions of log curves change abnormally at SBs and fs, being a good parameter in high-resolution sequence stratigraphic analysis and explanation of the complexity of geological background. Computer programs have been written in Matlab workspace to run Principal Component Analysis and wavelet denoising, providing an effective tool in corresponding study and cutting a new way to make full use of the well-logging data to solve geological problems in mineral exploration, especially within deep areas.