Particle Swarm Optimization with Family Communication Strategy
Particle swarm optimization (PSO) is a population-based stochastic algorithm for solv- ing complex optimization problems. To raise efficiency and accelerate convergence of PSO, we proposed a new sociological PSO algorithm with family concepts, named as FPSO. Here, family relationships and relative communication strategies were introduced into the conventional PSO algorithm. Two types of family relationships among parti- cles: equal relationship (ER) and generational relationship (GR) were introduced into the communication strategies among family members. The convergent speed and com- plexity of the proposed FPSO method were analyzed theoretically, and simulated by the IEEE-CEC 2015 learning-based benchmark problems to demonstrate the precision and convergent speed. And, the FPSO performances with ER and GR were separately tested and discussed. The experimental results indicated that the proposed FPSO method could improve the convergence performance, and had stronger judgment ability and intelligence than the conventional PSO method.
