Abstract
This paper proposes a new methodology for predicting aggregate flight departure delays in airports by exploring supervised learning methods. Individual flight data and meteorological information were processed to obtain four types of airport-related aggregate characteristics for prediction modeling. The expected departure delays in airports is selected as the prediction target while four popular supervised learning methods: multiple linear regression, a support vector machine, extremely randomized trees and LightGBM are investigated to improve the predictability and accuracy of the model. The proposed model is trained and validated using operational data from March 2017 to February 2018 for the Nanjing Lukou International Airport in China. The results show that for a 1-h forecast horizon, the LightGBM model provides the best result, giving a 0.8655 accuracy rate with a 6.65 min mean absolute error, which is 1.83 min less than results from previous research. The importance of aggregate characteristics and example validation are also studied.
Highlights
With the rapid development of civil aviation, flight delays have become an important subject and problem for air transportation systems all over the world
In China, the abnormal flights rate for 2018 was 80.13%, which means more than 850,000 flights were delayed during that year [2]. These flight delays have a severe economic impact in the U.S that is equivalent to 40.7 billion dollars per year [3], while a similar cost is expected for China
Passengers suffer a loss of time, missed business opportunities or leisure activities, and airlines attempting to make up for delays leads to extra fuel consumption and a larger adverse environmental impact
Summary
With the rapid development of civil aviation, flight delays have become an important subject and problem for air transportation systems all over the world. Many popular data driven methods have been used to predict flight delay, including the random forest algorithm, artificial neural network, logit probability and deep learning. Rodriguez et al [24] used an asymmetric logit probability model to estimate and predict the daily probabilities of delays in aircraft arrivals They identified that the origin-departure delay and distance between airports are significant delay factors distinct from the departure delay, the size of airline, the size of airport and the day of the flight. The main influence elements considered in these models are the time and weather-related factors, while some aggregate characteristics involving flight plans and airport delays have not yet been studied closely. The goal of this paper is to propose a methodology that can be used to predict flight departure delays in airports by exploring supervised learning methods considering aggregate flight data and local weather information in airports.
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
