The development of national economy and social productivity runs through all levels of people’s lives. The research and use of transportation tools are an inevitable trend of human civilization and development. The fundamental driving force lies in the pursuit of a better and convenient life. When every household is immersed in the pleasure of buying and using a car，the subsequent social impact cannot be ignored. The rapid increase in traffic flow has led to frequent road congestion. Road traffic is facing severe challenges on how to ensure safety and improve traffic efficiency at the same time；thus，the development of intelligent transportation systems cannot be delayed. In this paper，based on the deep learning method，the task of road speed prediction in the intelligent transportation system is studied，and a road speed prediction algorithm based on graph attention network is proposed. Firstly，the algorithm uses the recurrent neural network to learn the current road condition information，and then models the short-term and long-term characteristics of the neighbor road. Then the model uses the graph attention network to weigh the influence of the neighbor road to predict the speed of the target road. The effectiveness of the algorithm is verified in the real traffic dataset.