Research Article
LSTM-Based Prediction of Airport Aircraft in and Outflow
@INPROCEEDINGS{10.1007/978-3-030-90196-7_4, author={Baoqiang Li and Jin Huang and Yamei Duan and Yuan Zhao}, title={LSTM-Based Prediction of Airport Aircraft in and Outflow}, proceedings={Artificial Intelligence for Communications and Networks. Third EAI International Conference, AICON 2021, Xining, China, October 23--24, 2021, Proceedings, Part I}, proceedings_a={AICON}, year={2021}, month={11}, keywords={Airport traffic forecasting Long and short-term memory networks (LSTM) Support vector machines (SVM) Machine learning}, doi={10.1007/978-3-030-90196-7_4} }- Baoqiang Li
Jin Huang
Yamei Duan
Yuan Zhao
Year: 2021
LSTM-Based Prediction of Airport Aircraft in and Outflow
AICON
Springer
DOI: 10.1007/978-3-030-90196-7_4
Abstract
The prediction of airport inbound and outbound traffic is a hot research direction in civil aviation air traffic management. Using the historical data of air traffic as the data source of the traffic prediction model, the traffic data are processed and machine learning algorithm models such as Support Vector Machine (SVM) linear regression, LSTM (Long Short-Term Memory) recurrent neural network, and BP neural network are used to predict the air traffic. The experiments, analysis, and generalization of relevant machine learning algorithms for air traffic flow prediction are conducted. The experiments show that the prediction results are based on historical airspace traffic data, the LSTM model has the highest accuracy, the SVM linear regression has the second-highest prediction effect, and the BP neural network has a poor prediction effect and insufficient stability. The experimental results demonstrate that the LSTM-based inbound and outbound traffic prediction model can achieve airport traffic prediction based on historical traffic data. The usability and accuracy of the LSTM-based prediction results are illustrated by comparing different algorithms, proving that the LSTM model can be used for future urban air traffic flow prediction. It is demonstrated that the LSTM model can be used for future urban air traffic flow prediction. It provides a theoretical and reference basis for the future air traffic flow management of intelligent urban transportation systems.
