Drilling rate prediction using MLP neural network coupled with genetic algorithm (A case study: Maroon oil field)

Rate of penetration (ROP) is one of vital parameters and directly affects the drilling time and costs. There are various parameters that affect the drilling rate. Some of these parameters include weight on bit, rotational speed, mud weight, bit type, formation type and bit hydraulic. Reducing the total time of drilling as well as the drilling costs require an accurate prediction of ROP. In this study, an evolutionary algorithm called genetic algorithm is used to optimize the tuning parameters of MLP type of neural network (MLP-NN) to understand whether using this algorithm can be feasible in improving accuracy of MLP-NN for drilling rate prediction. The implemented smart system is trained using a large data set from Maroon oil field in Iran and its results are compared with a typical type of MLP-NN using different performance indices. Statistical and graphical analysis shows that typical type of neural network is more accurate than GA-MLP in prediction of drilling rate. It is found that employing genetic evolutionary algorithm to optimize the tuning parameters of MLP-NN cannot be an effective approach in order to improve prediction performance of these types of neural networks.

Drilling rate prediction using MLP neural network coupled with genetic algorithm (A case study: Maroon oil field) Keywords:

Rate of penetration , MLP neural network , Genetic algorithm , Maroon oil field