PREDIKSI PERFORMA SUMUR MINYAK MENGGUNAKAN ALGORITMA LONG SHORT-TERM MEMORY (LSTM) (STUDI KASUS : PT ELNUSA TBK)

Saputra, Rendi Cahya (2024) PREDIKSI PERFORMA SUMUR MINYAK MENGGUNAKAN ALGORITMA LONG SHORT-TERM MEMORY (LSTM) (STUDI KASUS : PT ELNUSA TBK). Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) cover.pdf Download (944kB)
	Text (BAB 1) 1.pdf Download (193kB)
	Text (BAB 2) 2.pdf Restricted to Repository staff only until 14 January 2028. Download (458kB)
	Text (BAB 3) 3.pdf Restricted to Repository staff only until 14 January 2028. Download (886kB)
	Text (BAB 4) 4.pdf Restricted to Repository staff only until 14 January 2028. Download (16MB)
	Text (BAB 5) 5.pdf Download (178kB)
	Text (Daftar Pustaka) dapus.pdf Download (177kB)
	Text (Lampiran) lampiran.pdf Restricted to Repository staff only until 14 January 2028. Download (1MB)

Abstract

Well testing is an important process in the petroleum industry because well testing serves to predict oil and gas production, which allows companies to make strategic decisions regarding production optimization and resource allocation. Conventional methods for well testing are time-consuming and costly. LSTM, a type of Recurrent Neural Network (RNN) with the addition of memory cells in order to store long-term information, is expected to overcome this weakness. This research uses time series data, which is used to study production patterns. The results showed that the LSTM model with the best scenario at choke 38 was obtained at 90% data split, hidden size 150 and 250, with MSE value 0.001 and RMSE value 0.044, for choke 40 was obtained at 90% data split, hidden size 150, with MSE value 0. 006 and RMSE value 0.078, and for choke 42 obtained at 90% data split, hidden size 150, with MSE value 0.007 and RMSE value 0.086, the results also show that the production results using choke 42 show the best performance compared to the other two chokes. This research shows that LSTM is an effective method for predicting well testing and can be an alternative for oil companies in optimizing production and making decisions on well production.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Anggraeny, Fetty Tri	NIP198202112021212005	fettyanggraeny.if@upnjatim.ac.id
Thesis advisor	Rizki, Agung Mustika	NIP199307252022031008	agung.mustika.if@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA75 Electronic computers. Computer science
T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Mahasiswa Rendi Saputra

Date Deposited:

14 Jan 2025 06:50

Last Modified:

15 Jan 2025 04:01

URI:

https://repository.upnjatim.ac.id/id/eprint/33891