Title | Application of computer neural network, and fuzzy set logic to petroleum geology, offshore eastern Canada |
Download | Download (whole publication) |
| |
Licence | Please note the adoption of the Open Government Licence - Canada
supersedes any previous licences. |
Author | Huang, Z; Shimeld, J; Williamson, M |
Source | Geological Survey of Canada, Current Research no. 1994-E, 1994 p. 243-250, https://doi.org/10.4095/194121 Open Access |
Year | 1994 |
Publisher | Natural Resources Canada |
Document | serial |
Lang. | English; French |
Media | paper; on-line; digital |
Related | This publication is contained in Current research 1994-E
|
File format | pdf |
Province | Eastern offshore region |
Area | Jeanne d'Arc Basin |
Subjects | mathematical and computational geology; analytical methods; computer applications; modelling; permeability; wells; fractures; structural features |
Illustrations | analyses |
Released | 1994 07 01 |
Abstract | Computer neural networks and fuzzy set logic are new advances in artificial intelligence that hold promise for geological applications. A neural network can model complicated geological problems better
than most conventional approaches. Fuzzy set logic can more efficiently process incomplete and imprecise information typically found in geosciences. In this paper, we outline both neural networks and fuzzy set logic and illustrate their utility in
geoscientific research. One example models the relationship between well logs and rock permeability through application of neural networks for the Venture Gas Field, Sable Basin. The trained neural network performs well in permeability prediction
from well logs. Another example demonstrates how open fracture are detected with conventional well logs using a fuzzy set logic algorithm in the Jeanne d'Arc Basin. The fractures index curves produced by fuzzy inference algorithms correlate with
known fractured zones, and reveal the extent of fracturing in zones where direct fracture information is not available. |
GEOSCAN ID | 194121 |
|
|