|
Control of gas injection and production oil flow of Rumaila oil field using FOPID controller and neural network
|
Farideh Allahverdi *1   , Jawad Kahdim Abduzahra Al-Dirawi2 |
1- Department of Electrical Engineering, Andimeshk Branch, Islamic Azad University, Andimeshk, Iran , a.allahverdi@iau.ac.ir
2- Department of Electrical Engineering, Dezful Branch, Islamic Azad University, Dezful, Iran |
|
|
Abstract: (66 Views) |
Today, various methods used to increase oil extraction in the world, which differ from each other according to the characteristics of each oil reservoir. In this research, FOPID control approach and neural network used in the direction of gas injection and production flow control of Rumaila oil field. In the neural network, the ten selected features that have the highest influence among the primary features include total volume, maximum well pressure, initial pressure, compressibility, pure to gross depth ratio, shear wave fraction, porosity to height ratio, shale to height ratio and permeability were noticed and used. The results by Matlab software showed that the MLP neural network with the training function trainbr and the number of neurons 10 and the number of layers 20 has an accuracy of 91.36% and the RBF neural network with SPREAD equal to 1, the number of neurons 26, and DF equal to 25 has an accuracy of 94.63% in gas injection control and They are the production discharge of the field. The MLP network with an accuracy of 97.46% and the RBF network with an accuracy of 98.97% optimize the gas injection and production flow rate of the field. |
|
| Keywords: flow rate control, oil field, neural network, FOPID controller. |
|
|
Full-Text [PDF 647 kb]
(22 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2025/05/24 | Accepted: 2025/08/20 | Published: 2025/09/1
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
