Cultural Algorithm Based on Particle Swarm Optimization for Function Optimization

A novel cultural algorithm based on particle swarm optimization was proposed in order to solve complex functions with high dimensions and overcome premature and the weak ability of local search. This algorithm model consists of population space and belief space, which have their own population evolve independently and parallelly. The lower level population space contributes elite individuals to the upper level belief space periodically, and in return the upper level belief apace evolves these elite individuals to influence the lower population space. Finally the dual evolution-dual improvement mechanism is established, which can improve the diversity of the population, get faster convergence speed, avoid premature problem, and obtain global optimum. Experimental results on several benchmark complex functions with high dimensions show that the proposed algorithm can rapidly converge at high equality solutions.