Particle Swarm algorithm as a new evolutionary optimization method can greatly reduce the computational burden of complex, large-scale optimization problems. This article is based on game theory. On the basis of the particle swarm it proposeda non-cooperative game model based on particle swarm optimization algorithm, it wses a game sequence repeated game model. And in repeated games, each game all hope to produce maximum benefits. Nash equilibrium of the corresponding game process. Function through multiple benchmarks, comparing with the performance of the algorithm experimental results show that the algorithm is feasible and effective. The study has important theoretical significance and practical significance onexpand swarm intelligence algorithm.