Abstract: In order to improve the energy utilization, reduce the carbon emission of transportation systems and promote the integration of integrated energy systems and transportation network in the background of building a new power system, the optimal dispatching model of traffic flow of fuel cell vehicles and hydrogen refueling behavior is constructed by considering the spatial and temporal distribution characteristics of load under the influence of hydrogen refueling behavior of fuel cell vehicles. Firstly, the origination-destination pair is used to portray the traffic flow model of fuel cell vehicles, and the Dijkstra algorithm is used to build the shortest hydrogen refueling path model for fuel cell vehicles to simulate the travel of fuel cell vehicle. Then, aiming at the uncertainties of renewable energy output, the fuzzy set of predicted errors for wind power is developed based on Wasserstein distance. And considering the constraints of power distribution networks and transportation networks, the distributionally robust optimization model considering the optimal dispatching of hydrogen refueling stations is developed. Finally, the distributionally robust optimization model is transformed into a mixed-integer linear programming model by using strong pairwise theory, and the scheduling method is validated by using the improved 33-bus distribution network and 24-bus traffic network. The results show that the total cost of the proposed method is reduced compared with the robust optimization system, and the scheduling scheme of hydrogen refueling stations achieves the balance between conservativeness and optimism, and realizes the positive interaction between hydrogen refueling stations and the higher-level network.