Abstract: Particle Swarm Optimization（PSO） is one of important Swarm Intelligence（SI） methods.The premature convergence and lower local search performance are drawbacks of PSO.This paper proposes a novel Gene Mutation PSO（GMPSO），some components of particles mutate according to the probability，a lot of experiments are taken.The results of the research and experiments indicate that the method can obviously improve the performance of PSO，it is credible in theory and feasible in technique.

Key words: swarm intelligence, particle swarm optimization, gene mutation

摘要： 粒子群优化（Particle Swarm Optimization，PSO）是一种重要的群智能（Swarm Intelligence，SI）方法。早期收敛和较低的局部搜索能力是PSO的不足。提出一种新颖的基因变异PSO（Gene Mutation PSO，GMPSO），依据概率使粒子的分量发生变异，并做了大量的实验。研究和实验的结果表明，该方法可显著改变PSO的性能，在理论上是可靠的，技术上是可行的。

关键词: 群智能, 粒子群优化, 基因变异