A Cooperative Particle Swarm Optimization Algorithm Based on Differential Evolution

This paper proposes a cooperative evolution particle swarm optimization(PSO) algorithm to preserve the variety of particle swarms and to avoid "premature" problem.The new algorithm uses two different particle swarms to search and find optimal value: sub-swarm one uses the standard PSO,and sub-swarm two uses the differential evolution algorithm.During the search,if the variety rate of fitness of the standard PSO is lower than a threshold,the bad particles in sub-swarm one are replaced with some good particles in sub-swarm two based on the principle of golden section.Then,the proposed PSO and the standard PSO are used to optimize four commonly-used functions.And the results show that the new PSO has higher efficiency,better performance and can find optimal value more easily than the standard PSO.