Integrated Intelligent Optimization based on Cultural Algorithm

This paper presents an Integrated Intelligent Optimization based on Cultural Algorithm (IIOBCA) to solve Optimization problems. Unlike the traditional CA, the IIOBCA consists of two belief spaces with different directions and a population space. The multi-algorithms are proposed to be used in the population space. There are two groups in the belief spaces, i.e. the “good” belief space and the “bad” belief space. From the population space, the “good” belief space only takes the individuals of the population with best fitness values, and the “bad” belief space takes the individuals with the rest “bad” fitness values. The evolution of the “good” belief space is adopted by Particle Swarm Optimization (PSO) and the “bad” belief space is adopted by Genetic Algorithm (GA) separately while the population are evolving at generation. The individuals in the population space are divided into three sub-groups, and the three groups are searching global optima via GA, PSO and COA respectively. To avoiding premature convergence, the strategy of exchanging the individuals between “good” belief space and “bad” belief space is presented. Regarding several large-dimension special unimodal and multimodal functions, compared with the basic GA, PSO and COA, simulation results show that the new algorithm can not only strengthen the global searching ability, but also overcome the disadvantage of falling into the local Optimization.