Permeability Prediction in Petroleum Reservoir using a Hybrid System

This paper introduces and demonstrates a hybrid soft computing system for predicting reservoir permeability of sedimentary rocks in drilled wells in the petroleum exploration and development industry. The method employs Takagi-Sugeno’s fuzzy reasoning, and its fuzzy rules and membership functions are automatically derived by neural networks and floating-point encoding genetic algorithms. The method is trained with known data and tested with unseen data. The results show that the hybrid system has a good generalisation capability and is effective for industrial applications.