«Proceedings of Gubkin University»

ARTIFICIAL NEURAL NETWORKS FOR SEPARATION OF SPACE TRENDS ON THE BASIS OF 3D SEISMIC RESULTS

UDC: 51.001
DOI: -

Authors:

¹ Gubkin Russian State University of Oil and Gas

Keywords: geological model, petrophysical properties, neural network

Annotation:

Interpolation of petrophysical properties is an important part of 3D geological modeling. 3D seismic contains huge amount of available information about acoustic parameters, which can (and must) be used for porosity prediction. In thes paper we suggest to use artificial neural networks for extracting such information from seismic volumes. This approach does not have significant drawbacks usual for some other model (for example, linear correlation).

Bibliography:

1. Aminzadeh F., P. de Groot. Neural Networks and Other soft computing Techniques with applications in the Oil industry. EAGE Publications, 2006, Netherlands, 164 p.
2. Осовский С. Нейронные сети для обработки информации. Пер. с польского.- М.: Финансы и статистика, 2004. - 344 с.
3. Урупов А.К. Основы трехмерной сейсморазведки. - М.: Изд-во Нефть и Газ, 2005.
4. Deutch, Clayton V. Geostatistical Reservoir modeling., Oxford University Press, New York, 2002.