A Hybrid GP-Fuzzy Approach for Resevoir Characterization

Abstract

A hybrid GP-fuzzy approach to model reservoirpermeability is presented. This approach uses atwo-step divide-and-conquer process for modelling.First, GP is applied to construct classifiers thatidentify permeability ranges. Within each range, ANFISis employed to build a Takagi-Sugeno-Kang fuzzyinference system that gives permeability estimation. Weapplied this method to five well log data sets. Theresults show that this hybrid system gives moreaccurate permeability estimation than other previousworks.