Fuzzy partitioning systems for electrofacies classification: A case study from the Maracaibo Basin

This paper describes a method of advanced data processing for the inverse p roblem of lithofacies prediction from well logs using fuzzy partitioning sy stems. A fuzzy partitioning system consists Of a set of fuzzy If-Then rules of the form "If bulk density (rho (b)) is low and neutron porosity (phi (C NL)) is high Then classify pattern x=(rho (b) phi (CNL) ) as Facies F-i". I n this paper, we introduce an intelligent method for the problem off fuzzy rule generation based on fuzzy clustering. Fuzzy clustering is used to dete ct structures in the multidimensional space of the available well log readi ngs. Each cluster detected is a potential fuzzy classification rule. By app lying fuzzy validity measures an optimum number of fuzzy clusters can be fo und. Using this approach, the number of rules, the antecedent membership fu nctions and other parameters that constitute the fuzzy partitioning system are derived in an automatic way. The aim is to find a minimum set of fuzzy classification rules that can cor rectly classify all log training patterns. Unlike traditional methods of pr edicting lithofacies, this approach does not require prior knowledge about the partitioning of the well log readings or any assumption of the facies p robability densities. Computer simulations using selected well log response s and facies description from a clastic and carbonate sequence in the Marac aibo Basin (western Venezuela) examine the performance of the fuzzy rule-ba sed classification approach. The performance of the fuzzy classification me thod is evaluated against the facies classification results using conventio nal statistical analysis.