With the increase of aviation activities, airspace constraints and flight delays have become increasingly prominent in the process of air traffic management (ATM). How to increase airspace capacity within the limited airspace resources while ensuring smooth and safe aircraft operations is a challenge for civil aviation today. In order to accelerate the intelligent operation of air traffic control and promote the development of intelligent ATM systems, it becomes very important to improve the accuracy of trajectory prediction, which becomes very difficult due to the problems of sparse aircraft flight trajectory and different flight altitudes. In order to solve this problem, an Attention-LSTM trajectory prediction model is proposed in this paper. This model can extract the time series features of the trajectory using an long short-term memory (LSTM) network and improve the accuracy of trajectory prediction by extracting the important factors affecting the current point change using an attention mechanism. To verify the accuracy of our proposed model, this paper uses real trajectory data, and uses LSTM, support vector machine, and other models for comparison. The comparison results show that this model is better than the most advanced model, which further promotes the process of intelligent ATM systems.