Correlative Particle Swarm Optimization Model
相关性粒子群优化模型

In the study of particle swarm optimization, propertly using the individual experience and social sharing information of particles has always been a problem. To solve this problem, this paper analyzes the random factors in updating the velocity eguation in the view of cognition and creates the intrinsic cognitive relation between individual experience and social sharing information. First, a correlative particle swarm optimization model is developed, which uses the Copula function to measure the dependence among random factors. In the new model, the different correlation structures and degrees of correlation between random factors can denote different strategies, which are used to process individual experience and social sharing experience. Meanwhile, this paper provides a flowchart of the correlative particle swarm optimization model, based on Gaussian Copula. Second, the relationship between the degrees of correlation and population diversity is presented, which shows that the random factors with positive linear correlation avail to maintain population diversity. Finally, the relationship between the degrees of correlation and convergence is analyzed and the convergence conditions of the correlative particle swarm optimization model are provided. Experimental simulations show that the correlation of random factors have a much greater influence on the performance of the new model, which can greatly improve convergence velocity and precision when the random factors are a completely positive linear correlation.