Jatilima : Jurnal Multimedia Dan Teknologi Informasi
Vol. 7 No. 02 (2025): Jatilima : Jurnal Multimedia Dan Teknologi Informasi

DEEP Q-NETWORK ANALYSIS IN OPTIMIZING DATA PROCESSING FOR DECISION MAKING ON FUEL EXPENDITURE FINANCE

Siregar, Andree Rizky Yuliansyah (Unknown)
Iqbal , Muhammad (Unknown)
Sitorus , Zulham (Unknown)
Novelan, Muhammad Syahputra (Unknown)
Darmeli Nasution (Unknown)



Article Info

Publish Date
04 Jun 2025

Abstract

As we know, Diesel fuel or also called Solar is a fuel used for diesel-engined motor vehicles, which are generally used in public transportation vehicles or commercial vehicles. In addition, it is also used in diesel for industry. Solar energy is obtained from petroleum refining. In addition to being a fuel, diesel also functions as a lubricant in diesel engine components. In managing the fuel budget for companies or agencies that have high operational needs, decisions regarding the allocation of funds and fuel purchases are very important. Inefficient or unplanned fuel purchases can result in waste and reduce profitability. Therefore, an optimal decision-making system is needed at PT. Deztonindo, which can accurately predict fuel needs and adjust the budget according to the company with the right price and market demand. This study uses a literature review method with the Deep Q-Network (DQN) method. The number of samples in this study is 2559 data with 10 test data. With a reduction in idle time of up to 50%, idle fuel consumption is reduced by 18 liters, increasing efficiency from 0.88 km / liter to 1.28 km / liter, or an increase of 45.5%. After optimization, there was a decrease in average fuel consumption of 20%, which had a direct impact on saving operational costs in a year for the diesel fuel purchase budget. The existence of this decision system can overcome the obstacles to obtaining accurate results for the diesel fuel purchase budget, to minimize the level of conditions that occur

Copyrights © 2025






Journal Info

Abbrev

jatilima

Publisher

Subject

Computer Science & IT

Description

JATILIMA merupakan jurnal yang terbit dua nomor dalam satu volume (tahun), yaitu Peridoe I Bulan April dan Periode II Bulan Oktober. JATILIMA mempublikasikan tulisan-tulisan ilmiah hasil pemikiran, studi literatur, dan penelitian dalam bidang Ilmu Komputer. JATILIMA merupakan jurnal dengan sistem ...