Taking into account the need of electric vehicles（EVs） of many public charging stations with their rapid growth in the future， public charging stations should be located according to the distribution of EVs. The method of predicting the distribution of EVs is given， and the charger configuration method based on the queuing theory is provided. A mathematical model for charging station location optimal planning is presented. The Voronoi diagram with many mathematical properties similar to EV charging stations is used to divide service areas. By considering the randomness of fast charging， the queuing time of EVs in the service areas calculated by the queuing theory M/M/s model is used to determine the scale of charging stations. To solve the Voronoi diagram global optimization problem， the Voronoi diagram is combined with the improved particle swarm optimization（PSO） algorithm， which has global stochastic optimization ability， and forms a joint solving process of the optimal planning model. Finally， the example analysis verifies the effectiveness and practicality of the planning method.