In air traffic flow management, air traffic flow forecasting is the most important part; it not only can improve security and effective utilization of airspace and airport resources but also can greatly improve airlines’ economy benefits and operational efficiency. In this paper, the Elman neural network prediction method, which achieved relatively good results in prediction of ground traffic, is applied in air traffic flow prediction, and a new air traffic flow prediction method based on multi-feature Elman neural network is proposed. First, the velocity characterized by the first derivative and the acceleration characterized by the second derivative are introduced as two new features into the structure of the single-feature Elman neural network, and a multi-feature Elman network is built. Further, the parameters of the network structure are studied by using the steepest descent method with the driving quantity items. The air traffic flow of 36 cities in East China are tested as the experimental data of the proposed method. Experimental results show that multi-feature Elman method compared with single-feature Elman method can obtain better prediction results.
Air Traffic Flow Forecasting Using Multi-Feature Elman Neural Network
22nd COTA International Conference of Transportation Professionals ; 2022 ; Changsha, Hunan Province, China
CICTP 2022 ; 137-147
2022-09-08
Conference paper
Electronic Resource
English
Short-term traffic flow prediction method based on GWO-Elman neural network
European Patent Office | 2023
|Researching of Chaotic Characteristics and Forecasting of Elman Network on the Communication Traffic
Trans Tech Publications | 2013
|UAV Photogrammetry and AFSA-Elman Neural Network in Slopes Displacement Monitoring and Forecasting
Springer Verlag | 2020
|Traffic Jam Prediction through Elman Neural Network Based on Monte Carlo Simulation
Trans Tech Publications | 2014
|