Abstract: A new approach is suggested to improve PSO’s performance called algorithm PSO-DSF.The dynamic Scale-Free like network is brought as the neighborhood topology structures and the mechanism of the dynamic directed network is designed.As the network evolves with the mechanism， the out degrees of the network follow the power-law distribution and the diversity of the algorithm is improved.The algorithm is identified that perform well when it is near optima positions.Four benchmark functions are selected as the tested functions.The experimental results illustrate the advantage of PSO-DSF.

Key words: Particle Swarm Optimization, directed network, scale-free like network

摘要： 该文提出了一种改进的PSO算法PSO-DSF。引进有向类无标度网作为粒子群寻优的拓扑结构，提出作为粒子邻域拓扑的有向网络动态变化机制，使有向网络在出度服从幂律分布的条件下动态变化，从而提高算法的多样性，避免过早陷入局部最优的情况。通过函数测试，证实了该改进方案的有效性。

关键词: 粒子群优化算法, 有向网络, 类无标度网