Primal-Dual Interior Point Method Particle Swarm Optimization (pdipmPSO) Algorithm

Emmanuel Gbenga Dada, E. Ramlan

Research output: Contribution to conferencePaper

Abstract

Particle Swarm Optimization (PSO) is a metaheuristic optimization algorithm that have been used to solve complex optimization problems that the traditional techniques finds very
difficult to solve. The Interior-Point Methods (IPMs) are efficient tools for solving nonlinear optimization problems. The IPMs having constrains that are active at the current point, are now believed to be the most robust algorithms for solving large-scale nonlinear optimization problems. Though they are very efficient, but they are still plagued with several challenges such as how to handle of nonconvexity, the procedure for making the barrier constraint up to date is cumbersome despite the existence of nonlinearities, and the need to ensure progress toward the solution. In order to overcome some of the shortcomings of the standard PSO such as premature convergence
and particles been trapped at the local minimal, we proposed the Primal-Dual Interior Point Particle Swarm Optimization (pdipmPSO) to surmount the shortcomings of the original PSO. We applied the
Primal Dual to each particle in a finite number of iterations, and feed the PSO with the output of the Primal Dual. We compared the performance of our new algorithm (pdipmPSO) with IPM and PSO
using 13 different benchmark functions. Optimization results reveal that pdipmPSO performs better than PSO and IPM. Our proposed algorithm is shown to have great capacity to prevent premature convergence, and the curse of particles being trapped in the local minimal which have characterised many variants of PSO.
Original languageEnglish
Pages94-101
Number of pages7
Publication statusPublished - 23 Mar 2015
Event3rd International Conference on Advances in Engineering Sciences and Applied Mathematics - London, United Kingdom
Duration: 23 Mar 201524 Mar 2015
http://iieng.org/allproceedings.php/31

Conference

Conference3rd International Conference on Advances in Engineering Sciences and Applied Mathematics
CountryUnited Kingdom
CityLondon
Period23/03/1524/03/15
Internet address

Keywords

  • Particle swarm optimisat
  • Interior Point Method
  • Primal-Dual
  • gbest and lbest
  • Unimodal functions
  • Multimodal functions

Fingerprint Dive into the research topics of 'Primal-Dual Interior Point Method Particle Swarm Optimization (pdipmPSO) Algorithm'. Together they form a unique fingerprint.

Cite this