Long-term prediction of the crude oil price using a new particle swarm optimization algorithm

Document Type : Original Article

Authors

1 Department of physics, Sirjan University of Technology, Sirjan, Iran.

2 Department of Computer Engineering, Faculty of Engineering, University of Guilan, Rasht, Iran

3 Department of Mechanical Engineering, Sirjan University of Technology

4 Department of Mechanical Engineering, University of Politecnico di Milano, Milan, Italy.

5 Department of Mechanical Engineering, Faculty of Engineering, University of Guilan, Rasht, Iran.

Abstract

Oil is one of the most precious source of energy for the world and has an important role in the global economy. Therefore, the long-term prediction of the crude oil price is an important issue in economy and industry especially in recent years. The purpose of this paper is introducing a new Particle Swarm Optimization (PSO) algorithm to forecast the oil prices. Indeed, the PSO is a population-based optimization method inspired by the flocking behavior of birds. Its original version suffers from tripping in local minima. Here, the PSO is enhanced utilizing a convergence operator, an adaptive inertia weight and linear acceleration coefficients. The numerical results of mathematical test functions, obtained by the proposed algorithm and other variants of the PSO elucidate that this new approach operates competently in terms of the convergence speed, global optimality and solution accuracy. Furthermore, the effective variables on the long-term crude oil price are regarded and utilized as input data to the algorithm. The objective function of the optimization process considered in this research study is the summation of the square of the difference between the actual and the predicted oil prices. Finally, the long-term crude oil prices are accurately forecasted by the proposed strategy which proves its reliability and competence.

Keywords

Main Subjects

Journal of Energy Management and Technology
Volume 5, Issue 1
March 2021
Pages 17-22

Files

History

  • Receive Date: 05 December 2019
  • Revise Date: 18 June 2020
  • Accept Date: 18 June 2020
  • First Publish Date: 01 March 2021

Share

How to cite

Statistics