关键词: 双种群交流, 粒子群优化, 投资组合模型

Abstract:  Considering that standard Particle Swarm Optimization（PSO） has the severe problem of being stuck in local optimums, this paper puts forward an improved particle swarm optimization with two particle swarms communicating with each other on the basis of velocity mutation, leading to the problem above to be resolved. In addition, taking into account the existence of transaction cost and the restraints of long sale, short sale, integral number of transactions and so on in China, this paper builds a portfolio selection model which totally reflects current status in our country. It applies the improved PSO above to deal with the model. The result comes out that the model is integrated and effective in our country, and also the PSO with two particle swarms communicating with each other is proper and efficient.

Key words: two particle swarms communicating with each other, Particle Swarm Optimization（PSO）, portfolio selection model