Short-term Airline Passenger Flow Prediction Based on the Attention Mechanism and Gated Recurrent Unit Model

The pressure of civil aviation traffic is increasing with the prosperity of economy. The accurate forecast of civil aviation flow not only effectively improves the operation efficiency of airlines, but also brings considerable profits to airlines. However, the existing aviation flow forecasting methods generally have the problem of poor forecasting accuracy. Inaccurate traffic prediction models not only fail to bring benefits, but also waste resources of airlines to a certain extent. Therefore, a high-precision forecast of aviation flow is necessary. On the basis of attention mechanism, a high-precision aviation flow model is constructed. First, the deep belief network is used to reduce the dimension of the data. Then, the Gated Recurrent Unit model is used to extract the time series features of the reduced dimension data. Finally, the attention mechanism is used to preserve the key features to achieve high-precision prediction. By analyzing historical data, the model which we proposed can accurately perceive the evolution process of civil aviation traffic and realize the high-precision prediction of short-term passenger flow. Experimental results show that the prediction accuracy of the model in this paper is significantly higher than other existing models, and the application of this model will bring considerable benefits to airlines.