Recently particle swarm optimization (PSO) has been achieved more attention, but the basic PSO is easy plunged into local optima. This paper proposed an improvement to basic PSO, named MPSO, in which mutation operator was added as genetic algorithm. The mutation operator can mutate the population, and then help the particles escape from local optima. Experiments on five benchmark functions showed the PSO with mutation operator not only enlarged particles' explore range, but also boosted the algorithm's convergence.