Abstract: This paper presents a variants of particle swarm optimizers named Experience Improving Particle Swarm Optimizer（EIPSO） in which a operator called Experience Improving （EI） is introduced.The EI operator initializes the part of the particle’s experence and gets the other experience.The new particle’s experience is selected from these two experiences according to their fitness.In each iteration of step，EI operator is performed with the probability of p，the evolution of particles is executed with the probability of 1-p.The new optimizer enables the diversity of the swarm to be preserved to discourage premature convergence.The result of experiments demonstrates the effect of EIPSO.

Key words: experience self-improve, particle swarm, optimizer

摘要： 经验自举粒子群优化算法（EIPSO）是在粒子群算法中引入经验自举（EI）搜索算子，该算子的作用就是将随机选择的粒子个体经验的局部重新初始化构成候选经验。根据候选经验和原经验的适应值确定个体的新经验。在粒子进化的每一代，以概率p来执行经验自举搜索，以概率1－p执行经验指导下的进化搜索。EI算子的引入使粒子的搜索范围和多样性得到保持，同时在粒子收敛后算法仍然具有一定的搜索能力。对比实验结果表明该EIPSO算法的良好的综合性能。

关键词: 经验自举, 粒子群, 优化