In this paper, the open queueing network model is proposed for solving the problem of public transportation in cities. The vertices of the networks (i.e., the bus stops) are determined by means of the fuzzy clustering method. The arcs (i.e., the paths of the public transportation) can be set up by using the shortest path model in the time sense or the 0–1 integer programming method. Applying the statistics method, we can calculate the parameters (such as the passenger flow’s distribution, passenger flow’s transition probability, mean waiting time for the bus etc.) of the public transportation network. In this paper, we suggest to divide the network into two or three stages to implement the public transportation system in the form of “frog jumping” fast transfer and “permeation” fast dispersion. Combining the computer simulation and the evaluation of the achievement and effect of public transportation system, we modify the model so as to solve the public transportation problem better.