A PSO Algorithm based on Biologe Population Multiplication (PMPSO)

Lei Yin1 and Xiaoxiang Liu2

1School of Mechano-Electronic Engineering, Xidian University, Xi’an, China.

2Department of Computer Science of Zhuhai College, Jinan University, Zhuhai, China.


Inspired by the natural phenomenon of multiplication of biological population, a population multiplication particle swarm optimization (PMPSO) is presented. The proposed algorithm (PMPSO) has four phases of migration, selection, elimination and reproduction, evolution. Using searching optimal model of PSO in the migration phase; introducing LEVEL SET theory dividing population to be able to facilitate the selection operation in the selection phase; speeding up the algorithm convergence by abandoning the inferior population, reproducing superior population and making full use of population resource in the phase of elimination and reproduction; creating new population to keep the diversity to avoid monotone of the algorithm in the last evolutionary phase. Finally, PMPSO is applied to some test functions comparing with GA and SPSO algorithm, which is proved that the PMPSO is feasible and effective.

Keywords: Biologe population multiplication, LEVEL SET, Particle swarm optimization.

     Back to TOC