中文摘要：

将共享机制引入微粒群算法，把群体的粒子适应度更新为共享适应度，对共享适应度高的粒子进行处罚，保留低适应值的粒子为记忆粒子，当全局最好值连续进化若干代无变化时，用记忆粒子和克隆选择来更新粒子。这样既增加了群体的多样性，同时又保存了群体中最好的粒子，从而有效克服了由于微粒群算法多样性差而造成的易陷于局部最优和对多峰值函数搜索效果不佳的缺点，仿真实验验证了该算法的有效性。

英文摘要：

Sharing mechanism was introduced into the particle swarm optimization. Particle＇s fitness value of population was updated for sharing fitness value. Particles with higher share fitness value were punished and particles with smaller sharing fitness value were remained as memory particles. Particles were updated with memory particles and clone choice when global best did not changed in some continuous evolutions. In this way, diversity of population is increased. At the same time the particle that has the best fitness value is saved. The modified algorithm can avoid the local optimization and has better search performance for multi-peak functions. The experimental results show the modified algorithm has better convergence performance than original particle swarm optimization algorithm.