Application of Neurofuzzy pattern recognition method in borehole geophysics

Geophysical data are specific physical responses of geological formations distributed over an area. These data are normally the physical parameters such as density, velocity, resistivity, susceptibility etc. of geological sources and hence bring a pattern of geological structures. It is conceived ac...

Full description

Bibliographic Details
Published in:Acta Geodaetica et Geophysica Hungarica
Main Authors: Singh, U., Singh, D., Singh, H.
Format: Article in Journal/Newspaper
Language:Hungarian
Published: Akadémiai Kiadó 2010
Subjects:
Online Access:http://real.mtak.hu/82260/
http://real.mtak.hu/82260/1/ageod.45.2010.4.2.pdf
https://doi.org/10.1556/AGeod.45.2010.4.2
Description
Summary:Geophysical data are specific physical responses of geological formations distributed over an area. These data are normally the physical parameters such as density, velocity, resistivity, susceptibility etc. of geological sources and hence bring a pattern of geological structures. It is conceived accordingly that this pattern recognition of such geophysical data will throw light on the spatial distribution and physical attributes of their geological sources. The well logging method considered as one of the geophysical method for the exploration of hydrocarbon, coal and base-metals, also has a strong role in finding the location and evaluation of geological resources.A novel approach known as Adaptive Neurofuzzy Inference System technique (ANFIS) is made to identify stratigraphy of Prydz Bay basin, east Antarctica. A geological stratum in terms of 1D model is made using datasets obtained from this area. The 1D model deduced as an ANFIS result is able to make geological sense of even additional thin sand sandwiched between clayey silt strata seams which unable to be resolved by other conventional methods. The analysed ANFIS results deduced to map horizons for hydrocarbon prospecting is verified with known coring datasets. These results are encouraging and provide stable and consistent solutions.