An effective and efficient two stage algorithm for global optimization

  • Authors:

  • Yong-Jun Wang;Jiang-She Zhang;Yu-Fen Zhang

  • Affiliations:

  • Institute of Information and System Science, School of Science, Xi’an Jiaotong University, Xi’an, China;Institute of Information and System Science, School of Science, Xi’an Jiaotong University, Xi’an, China;Faculty of Mathematics and Computer Science, Hebei University, Baoding, China

  • Venue:
  • ICMLC'05 Proceedings of the 4th international conference on Advances in Machine Learning and Cybernetics
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

A two stage algorithm, consisting of gradient technique and particle swarm optimization (PSO) method for global optimization is proposed. The gradient method is used to find a local minimum of objective function efficiently, and PSO with potential parallel search is employed to help the minimization sequence to escape from the previously converged local minima to a better point which is then given to the gradient method as a starting point to start a new local search. The above search procedure is applied repeatedly until a global minimum of the objective function is found. In addition, a repulsion technique and partially initializing population method are also incorporated in the new algorithm to increase its global search ability. Global convergence is proven, and tests on benchmark problems show that the proposed method is more effective and reliable than the existing optimization methods.