Comparison of WAVENET and ANN for predicting the porosity obtained from well log data

Porosity is one of the most important parameters of the hydrocarbon reservoirs, the accurate knowledge of which allows petroleum engineers to have adequate tools to evaluate and minimize the risk and uncertainty in the exploration and production of oil and gas reservoirs. Different direct and indirect methods are used to measure this parameter, most of which (e.g. core analysis) are very time-consuming as well as cost-consuming. Hence, applying an efficient method that can model this important parameter is of the highest importance. Most of the researches show that the capability (i.e. classification, pattern matching, optimization and data mining) of an ANN is suitable for inherenting uncertainties and imperfections found in petroleum engineering problems considering its successful application. In this paper, an alternative method of porosity prediction, which is based on integration between wavelet theory and Artificial Neural Network (ANN) or wavelet network (wavenet), is presented. In this study, different wavelets are applied as activation functions to predict the porosity from well log data. The efficacy of this type of network in function learning and estimation is compared with ANNs. The simulation results indicate decrease in estimation error values that depicts its ability to enhance the function approximation capability and consequently exhibits excellent learning ability compared to the conventional neural network with sigmoid or other activation functions.