eprintid: 10527
rev_number: 2
eprint_status: archive
userid: 1
dir: disk0/00/01/05/27
datestamp: 2023-11-09 16:37:08
lastmod: 2023-11-09 16:37:08
status_changed: 2023-11-09 16:31:37
type: article
metadata_visibility: show
creators_name: Momeni, M.
creators_name: Hosseini, S.J.
creators_name: Ridha, S.
creators_name: Laruccia, M.B.
creators_name: Liu, X.
title: An optimum drill bit selection technique using artificial neural networks and genetic algorithms to increase the rate of penetration
ispublished: pub
note: cited By 16
abstract: Drill bit is the most essential tool in drilling and drill bit selection plays a significant role in drilling process planning. This paper discusses bit selection by employing a method of combining Artificial Neural Network (ANN) and the computation of Genetic Algorithm (GA). In this method, offset well drilling records are used for training the ANN model and International Association Drilling Contractors (IADC) bit codes are used to name each bit. However, some researchers have used bit codes as input or output variables. This paper illustrates that the bit codes are better used in referring to the name of each bit instead of using them as values for calculation in the ANN model. The ANN black box was converted to white box to obtain a visible mathematical model for predicting the Rate of Penetration (ROP). This mathematical model, which was the objective function in the GAs, was used to find the optimum drilling values and to maximize the ROP. When drilling a new well, bit selection process requires the maximum ROP of a bit that corresponds to the optimum drilling parameters being obtained by combining the trained ANN model with GA. A bit selection example is provided by using the Shadegan oil field drilling data. The mean square error (MSE) obtained a value of 0.0037 whereas the coefficient of determination obtained a value of 0.9473. In other words, the predicted ROP model based on the field drilling data indicated a good correlation with the real ROP. © School of Engineering, Taylor�s University.
date: 2018
publisher: Taylor's University
official_url: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85041802992&partnerID=40&md5=433c9205b717f02473875d0f19d8d7cf
full_text_status: none
publication: Journal of Engineering Science and Technology
volume: 13
number: 2
pagerange: 361-372
refereed: TRUE
issn: 18234690
citation:   Momeni, M. and Hosseini, S.J. and Ridha, S. and Laruccia, M.B. and Liu, X.  (2018) An optimum drill bit selection technique using artificial neural networks and genetic algorithms to increase the rate of penetration.  Journal of Engineering Science and Technology, 13 (2).  pp. 361-372.  ISSN 18234690