On the utilisation of fuzzy rule-based systems for taxi time estimations at airports

Chen, Jun and Ravizza, Stefan and Atkin, Jason and Stewart, Paul (2011) On the utilisation of fuzzy rule-based systems for taxi time estimations at airports. In: 11th Workshop on Algorithmic Approaches for Transportation Modelling, Optimization, and Systems, 8th September 2011, Saarbrucken, Germany.

Full text not available from this repository.

Official URL: http://drops.dagstuhl.de/opus/volltexte/2011/3273/

Abstract

The primary objective of this paper is to introduce Fuzzy Rule-Based Systems (FRBSs) as a relatively new technology into airport transportation research, with a special emphasis on ground movement operations. Hence, a Mamdani FRBS with the capability to learn from data has been adopted for taxi time estimations at Zurich Airport (ZRH). Linear regression is currently the dominating technique for such an estimation task due to its established nature, proven mathematical characteristics and straightforward explanatory ability. In this study, we demonstrate that FRBSs, although having a more complex structure, can offer more accurate estimations due to their proven properties as nonlinear universal approximators. Furthermore, such improvements in accuracy do not come at the cost of the model's interpretability. FRBSs can offer more explanations of the underlying behavior in different regions. Preliminary results on data for ZRH suggest that FRBSs are a valuable alternative to already established linear regression methods. FRBSs have great potential to be further seamlessly integrated into the taxiway routing and scheduling process due to the fact that more information is now available in the explanatory variable space.

Item Type:Conference or Workshop Item (Presentation)
Additional Information:The primary objective of this paper is to introduce Fuzzy Rule-Based Systems (FRBSs) as a relatively new technology into airport transportation research, with a special emphasis on ground movement operations. Hence, a Mamdani FRBS with the capability to learn from data has been adopted for taxi time estimations at Zurich Airport (ZRH). Linear regression is currently the dominating technique for such an estimation task due to its established nature, proven mathematical characteristics and straightforward explanatory ability. In this study, we demonstrate that FRBSs, although having a more complex structure, can offer more accurate estimations due to their proven properties as nonlinear universal approximators. Furthermore, such improvements in accuracy do not come at the cost of the model's interpretability. FRBSs can offer more explanations of the underlying behavior in different regions. Preliminary results on data for ZRH suggest that FRBSs are a valuable alternative to already established linear regression methods. FRBSs have great potential to be further seamlessly integrated into the taxiway routing and scheduling process due to the fact that more information is now available in the explanatory variable space.
Keywords:Fuzzy Rule-Based Systems, Aircraft taxiing
Subjects:G Mathematical and Computer Sciences > G700 Artificial Intelligence
H Engineering > H411 Air-Passenger Transport Engineering
Divisions:College of Science > School of Engineering
ID Code:6057
Deposited By:INVALID USER
Deposited On:18 Aug 2012 19:28
Last Modified:18 Aug 2012 19:28

Repository Staff Only: item control page