International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 184 - Issue 52 |
Published: Mar 2023 |
Authors: Nlerum Promise Anebo, Igbudu Kingsley Ezebunwo |
![]() |
Nlerum Promise Anebo, Igbudu Kingsley Ezebunwo . A Comparative Study of an AI Pipeline Monitoring System and Particle Swarm Optimization Technique in Predictive Monitoring Operations: Oil and Gas Pipeline Vandalism. International Journal of Computer Applications. 184, 52 (Mar 2023), 39-49. DOI=10.5120/ijca2023922650
@article{ 10.5120/ijca2023922650, author = { Nlerum Promise Anebo,Igbudu Kingsley Ezebunwo }, title = { A Comparative Study of an AI Pipeline Monitoring System and Particle Swarm Optimization Technique in Predictive Monitoring Operations: Oil and Gas Pipeline Vandalism }, journal = { International Journal of Computer Applications }, year = { 2023 }, volume = { 184 }, number = { 52 }, pages = { 39-49 }, doi = { 10.5120/ijca2023922650 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2023 %A Nlerum Promise Anebo %A Igbudu Kingsley Ezebunwo %T A Comparative Study of an AI Pipeline Monitoring System and Particle Swarm Optimization Technique in Predictive Monitoring Operations: Oil and Gas Pipeline Vandalism%T %J International Journal of Computer Applications %V 184 %N 52 %P 39-49 %R 10.5120/ijca2023922650 %I Foundation of Computer Science (FCS), NY, USA
This work proposed an improved neural network model known as AI Pipeline Monitoring System for Predictive Monitoring of oil and gas installation vandalism threats. The system employed a sparse representative long-short-memory (SLSTM) learning network as part of a refinement to an existing feed-forward neural network. The system also uses a Gaussian membership function with a context-decision gate for detection and monitoring operations. In this paper the proposed system's efficiency is compared to that of the Particle Swarm Optimization Technique; a swarm intelligence algorithm that is emerging as an alternative to more conventional approaches for predictive monitoring operations. To test and evaluate the performance, dynamic simulations were performed using real-time dataset of most likely vandal behavior and the efficiency of the two systems in predictive monitoring operations. The results of simulation study showed impressive results and proves that the AI Pipeline Monitoring System is more preferred to the Particle Swarm System, because of its (AI Pipeline Monitoring System) continual long range context learning capability, which is a likely feature of most observed pipeline threat context-data.