Full text loading...
-
Application of a New Machine Learning Method on a Cascade of Fuzzy-Logical Matrices for Petroleum Reservoir Geological and Hydrodynamic Simulation
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, Saint Petersburg 2020, Nov 2020, Volume 2020, p.1 - 6
Abstract
The traditional version of geological and hydrodynamic models is not at all the only possible one, and, moreover, it impedes the development of geological and hydrodynamic modeling of oil and gas fields as a whole, since it is not truly mathematical. Considering the fact that computers do not work with images, but with numbers, the idea arose to develop a new type of geological and hydrodynamic models that have an unusual appearance, but they are more effective for computer analysis and forecasting. Geological and hydrodynamic models of a new type are a cascade of several tens or hundreds of fuzzy-logical matrices. In particular, such an approach can be effective for oil and gas fields with unconventional reserves, which are characterized by serious limitations on the quantity and quality of the source information.