Suitability Determination of Machine Learning Techniques for the Operational Quality Assessment of Geophysical Survey Results
2020
Kirill Abramov, Jānis Grundspeņķis

Well logging, also known as a geophysical survey, is one of the main components of a nuclear fuel cycle. This survey follows directly after the drilling process, and the operational quality assessment of its results is a very serious problem. Any mistake in this survey can lead to the culling of the whole well. This paper examines the feasibility of applying machine learning techniques to quickly assess the well logging quality results. The studies were carried out by a reference well modelling for the selected uranium deposit of the Republic of Kazakhstan and further comparing it with the results of geophysical surveys recorded earlier. The parameters of the geophysical methods and the comparison rules for them were formulated after the reference well modelling process. The classification trees and the artificial neural networks were used during the research process and the results obtained for both methods were compared with each other. The results of this paper may be useful to the enterprises engaged in the geophysical well surveys and data processing obtained during the logging process.


Atslēgas vārdi
Classification trees, machine learning, neural networks, well logging
DOI
10.2478/acss-2020-0017

Abramov, K., Grundspeņķis, J. Suitability Determination of Machine Learning Techniques for the Operational Quality Assessment of Geophysical Survey Results. Applied Computer Systems, 2020, Vol. 25, No. 2, 153.-162. lpp. ISSN 2255-8683. e-ISSN 2255-8691. Pieejams: doi:10.2478/acss-2020-0017

Publikācijas valoda
English (en)
RTU Zinātniskā bibliotēka.
E-pasts: uzzinas@rtu.lv; Tālr: +371 28399196