Cooperative micro-particle swarm optimization
Konstantinos E. Parsopoulos · 2009
Cooperative approaches have proved to be very useful in evolutionary computation due to their ability to solve efficiently high-dimensional complex problems through the cooperation of low–dimensional subpopulations. On the other hand, Micro–evolutionary approaches employ very small populations of just a few individuals to provide solutions rapidly. However, the small population size renders them prone to search stagnation. This paper introduces Cooperative Micro– Particle Swarm Optimization, which employs cooperative low–dimensional and low–cardinality subswarms to concurrently adapt different subcomponents of high–dimensional optimization problems. The algorithm is applied on highdimensional instances of five widely used test problems with very promising results. Comparisons with the standard Particle Swarm Optimization algorithm are also reported and discussed.