Please use this identifier to cite or link to this item:
|Title:||Taxi-speed prediction by spatio-temporal graph-based trajectory representation and its applications||Authors:||Tran, Thanh-Nam
|Keywords:||Engineering::Aeronautical engineering::Air navigation||Issue Date:||2020||Source:||Tran, T., Pham, D., Alam, S. & Duong, V. (2020). Taxi-speed prediction by spatio-temporal graph-based trajectory representation and its applications. International Conference for Research in Air Transportation (ICRAT) 2020.||Abstract:||Airport surface movement systems require aircraft taxing speed as a key input to perform ground movement optimization and path planning processes. With the increasing availability of surface movement data from systems such as A-SMGCS, a data-driven framework using a spatio-temporal graph-based trajectory representation is proposed in this paper to predict aircraft taxing speed. The proposed framework includes a data preparation module for converting track points data to graph-based representation and a developing predictive model module for learning taxi-speed model. The Random Forest algorithm is selected as our predictive model. The model predicts the aircraft taxi-speed with an error of 1.08 m/s for taxi-out procedure and 0.97 m/s for taxi-in procedure, when compared with the actual taxi-speed from A-SMGCS data, respectively. Further, three applications of our approach are discussed which are taxi-speed profile, unimpeded taxi time and potential conflict detection. The results of our methods outperform all baseline methods. In detail, for generating taxi-speed profile, our method obtains the error 1.38 m/s while for computing unimpeded taxi time, our method outperforms the baseline model with the mean absolute percentage error is 11.03% for the taxi-in and 16.8% for taxi-out procedure, respectively.||URI:||https://hdl.handle.net/10356/146853||Rights:||© 2020 ICRAT. All rights reserved. This paper was published in International Conference for Research in Air Transportation (ICRAT) and is made available with permission of ICRAT.||Fulltext Permission:||open||Fulltext Availability:||With Fulltext|
|Appears in Collections:||ATMRI Conference Papers|
Updated on Jun 27, 2022
Updated on Jun 27, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.