Predicting fuel consumption has become an important research topic to reduce the adverse impact of fuel emissions on the environment, save fuel energy resources, reduce flight costs, achieve more accurate aircraft trajectory prediction, and provide efficient air traffic management. This paper proposed a deep learning model to predict fuel consumption of civil aircraft. The experimental data is collected from the aircraft flight recorder in the form of QAR data, and the prediction results reach 90% accuracy. Evaluation results are applied to the problem of energy consumption management in supply chain operation.