TY - GEN
T1 - Piping system development in oil and gas field based on Dijkstra's algorithm and genetic algorithm
AU - Gunawan,
AU - Utomo, Allessandro Setyo Anggito
AU - Syafaq, Altabibi Naufal
N1 - Publisher Copyright:
© 2021 Author(s).
PY - 2021/9/23
Y1 - 2021/9/23
N2 - Modelling of pipeline routes in the oil and gas industry is necessary to examine in this decade. The complexity of the pipeline design in the oil and gas business increases operational costs over time. In addition to operational costs, the problem with pipe design is that humans cannot calculate with certainty the physical limitations that occur in the pipeline. Therefore, produce a level of certainty in the pipeline design is fallacious. The use of Genetic Algorithms (GA) and Dijkstra's Algorithm (DA) as artificial intelligence produce easy calculation and prediction of pipelines to the optimal and accurate stages. The pipeline simulation is based on the pipe's length and the amount of bending as a physical attribute to pressure drop in pipeline. In this study, the algorithm uses 2000 data generation from 100 populations, 80% of the data used for training the algorithm model (1600 experiments), and 20% (400 experiments) of data used for control-the experiment conducted in a 1: 5000 (65 x 60 grid) environment. The research used a map of the distribution of oil and gas pipelines in Nigeria around the regions of Benin, Warri, Owerri, Aba, and Harcourt. Pipeline optimization from genetic algorithm modelling and Dijkstra's Algorithm increased prediction efficiency by 4.9%.
AB - Modelling of pipeline routes in the oil and gas industry is necessary to examine in this decade. The complexity of the pipeline design in the oil and gas business increases operational costs over time. In addition to operational costs, the problem with pipe design is that humans cannot calculate with certainty the physical limitations that occur in the pipeline. Therefore, produce a level of certainty in the pipeline design is fallacious. The use of Genetic Algorithms (GA) and Dijkstra's Algorithm (DA) as artificial intelligence produce easy calculation and prediction of pipelines to the optimal and accurate stages. The pipeline simulation is based on the pipe's length and the amount of bending as a physical attribute to pressure drop in pipeline. In this study, the algorithm uses 2000 data generation from 100 populations, 80% of the data used for training the algorithm model (1600 experiments), and 20% (400 experiments) of data used for control-the experiment conducted in a 1: 5000 (65 x 60 grid) environment. The research used a map of the distribution of oil and gas pipelines in Nigeria around the regions of Benin, Warri, Owerri, Aba, and Harcourt. Pipeline optimization from genetic algorithm modelling and Dijkstra's Algorithm increased prediction efficiency by 4.9%.
KW - Algorithm
KW - Artificial intelligence
KW - Optimization
KW - Pipeline routes
KW - Prediction
UR - http://www.scopus.com/inward/record.url?scp=85116501211&partnerID=8YFLogxK
U2 - 10.1063/5.0063889
DO - 10.1063/5.0063889
M3 - Conference contribution
AN - SCOPUS:85116501211
T3 - AIP Conference Proceedings
BT - 5th International Tropical Renewable Energy Conference, i-TREC 2020
A2 - Irwansyah, Ridho
A2 - Budiyanto, Muhammad Arif
PB - American Institute of Physics Inc.
T2 - 5th International Tropical Renewable Energy Conference, i-TREC 2020
Y2 - 29 October 2020 through 30 October 2020
ER -