An integrated neural-fuzzy-genetic-algorithm using hyper-surface membership functions to predict permeability in petroleum reservoirs

This paper introduces a new neural-fuzzy technique combined with genetic al gorithms in the prediction of permeability in petroleum reservoirs. The met hodology involves the use of neural networks to generate membership functio ns and to approximate permeability automatically From digitized data (well logs) obtained from oil wells. The trained networks are used as fuzzy rules and hyper-surface membership Functions. The results of these rules are int erpolated based on the membership grades and the parameters in the defuzzif ication operators which are optimized by genetic algorithms. The use of the integrated methodology is demonstrated via a case study in a petroleum res ervoir in offshore Western Australia, The results show that the integrated neural-fuzzy-generic-algorithm (INFUGA) gives the smallest error on the uns een data when compared to similar algorithms. The INFUGA algorithm is expec ted to provide a significant improvement when the unseen data come from a m ixed or complex distribution. (C) 2001 Elsevier Science Ltd. All rights res erved.