Applications of Memory Based Particle Swarm Optimization for Predicting Equipment Failures and Optimizing Cost in Oil and Gas

  • Chaitanya K., Somayajulu D. V. L. N., P. Radha Krishna

Abstract

Operations in the oil and gas industry, such as oil extraction and oil transportation, require complex machinery systems. The oil transportation system consists of thousands of kilometers of hydraulic pipelines to transport the extracted oil to the place of interest. Maintaining these large systems is not an easy task. Oil and gas companies invest millions of dollars in monitoringand maintaining the pipelines to ensure that they run 24/7. Researchers have proposed various models for predicting such failures. This paper presents two approaches for predicting equipment failures. The two approaches for equipment failure are detected by an algorithm called Memory-based Particle Swarm Optimization (MS-PSOMR) based on equipment sensors' values. The goal of the proposed methods is to store the objective function values in the particle's memory and use them to detectthe failure. In the first technique, we model the equipment failure detection as a constrained optimization problem and solve it by minimizing the difference between actual and observed sensor values. In the second technique, we calculate the equipment's failure rate and store them in particle memory for predicting the equipment failure. The proposed technique is on public datasets to validate the efficiency of the proposed techniques.        

Published
2021-05-17
How to Cite
Chaitanya K., Somayajulu D. V. L. N., P. Radha Krishna. (2021). Applications of Memory Based Particle Swarm Optimization for Predicting Equipment Failures and Optimizing Cost in Oil and Gas. Design Engineering, 664 - 682. Retrieved from http://www.thedesignengineering.com/index.php/DE/article/view/1583
Section
Articles