Modeling the Permeability of a Reservoir Using a New Hybrid Fuzzy-Neural Network Strategy

Abstract

Knowledge of permeability is critical to developing an effective reservoir description. Permeability data can be obtained directly from well tests and cores or indirectly from logs. Normally, using well log data to derive estimates of permeability is the lowest cost method. This article proposes a structured multiple neural network that is constructed according to the structure of IF-THEN fuzzy inference rules and the fuzzy c-means clustering method, which combines the advantages of neural network and fuzzy systems. This multiple neural network shows that the predictive performance of the proposed model is superior to that of both the fuzzy logic and neural network.