Particle Swarm Optimization with Inertia Weight and Constriction Factor

In the original Particle Swarm Optimization (PSO) formulation, convergence of a particle towards its attractors is not guaranteed. A velocity constraint is successful in controlling the explosion, but not in improving the fine-grain search. Clerc and Kennedy studied this system, and proposed constriction methodologies to ensure convergence and to fine tune the search. Thus, they developed different constriction methods according to the correlations among some coefficients incorporated to the system. Type 1” constriction became very popular because the basic update equations remained virtually unmodified, and the original and intuitive metaphor valid. The main drawbacks of this constriction type are that constriction becomes too strong very quickly as the acceleration is increased; that the speed of convergence cannot be easily controlled; and that there is no flexibility to set a desired form of convergence. Another problem is that the method can be found in the literature formulated so as to constrict a PSO system which already includes the inertia weight, for which the calculation of a constriction factor using the formulae provided by Clerc and Kennedy does not guarantee convergence. This paper analyzes Type 1” constriction in detail, for which Type 1 constriction is also relevant. The formulae for Type 1 and Type 1” constriction factors suitable for a PSO algorithm including the inertia weight are provided.

[1]  Hussein A. Abbass,et al.  A Modified Strategy for the Constriction Factor in Particle Swarm Optimization , 2007, ACAL.

[2]  Shangxu Hu,et al.  A New Approach to Improve Particle Swarm Optimization , 2003, GECCO.

[3]  Michael N. Vrahatis,et al.  Tuning PSO Parameters Through Sensitivity Analysis , 2002 .

[4]  Maurice Clerc,et al.  The particle swarm - explosion, stability, and convergence in a multidimensional complex space , 2002, IEEE Trans. Evol. Comput..

[5]  Johann Sienz,et al.  ICSI 2011 : International conference on swarm intelligence id – 1 Memetic Particle Swarm for Continuous Unconstrained and Constrained Optimization Problems , 2011 .

[6]  Mauro Sebastián Innocente.  Development and testing of a particle swarm optimizer to handle hard unconstrained and constrained problems , 2010 .

[7]  Hitoshi Iba,et al.  Particle swarm optimization with Gaussian mutation , 2003, Proceedings of the 2003 IEEE Swarm Intelligence Symposium. SIS'03 (Cat. No.03EX706).

[8]  Ioan Cristian Trelea,et al.  The particle swarm optimization algorithm: convergence analysis and parameter selection , 2003, Inf. Process. Lett..

[9]  James Kennedy,et al.  The Behavior of Particles , 1998, Evolutionary Programming.

[10]  James Kennedy,et al.  Particle swarm optimization , 2002, Proceedings of ICNN'95 - International Conference on Neural Networks.

[11]  E. Ozcan,et al.  Particle swarm optimization: surfing the waves , 1999, Proceedings of the 1999 Congress on Evolutionary Computation-CEC99 (Cat. No. 99TH8406).

[12]  Chilukuri K. Mohan,et al.  Analysis of a simple particle swarm optimization system , 1998 .

[13]  Yue Shi,et al.  A modified particle swarm optimizer , 1998, 1998 IEEE International Conference on Evolutionary Computation Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98TH8360).

[14]  Frans van den Bergh,et al.  An analysis of particle swarm optimizers , 2002 .

[15]  Vishnu Prasad,et al.  Economic dispatch using particle swarm optimization: A review , 2009 .

[16]  Andries Petrus Engelbrecht,et al.  A study of particle swarm optimization particle trajectories , 2006, Inf. Sci..

[17]  Wei-Ping Lee,et al.  Adaptive constriction factor for location-related particle swarm , 2007 .

[18]  James Kenedy.  How It Works: Collaborative Trial and Error , 2008 .