Trajectory prediction uncertainty modelling for Air Traffic Management

Casado Magaña, Enrique Juan (2016) Trajectory prediction uncertainty modelling for Air Traffic Management. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2016CasadoPhD.pdf] PDF
Download (6MB)
Printed Thesis Information: https://eleanor.lib.gla.ac.uk/record=b3255208

Abstract

The anticipated growth of air traffic worldwide requires enhanced Air Traffic Management (ATM) technologies and procedures to increase the system capacity, efficiency, and resilience, while reducing environmental impact and maintaining operational safety. To deal with these challenges, new automation and information exchange capabilities are being developed through different modernisation initiatives toward a new global operational concept called Trajectory Based Operations (TBO), in which aircraft trajectory information becomes the cornerstone of advanced ATM applications. This transformation will lead to higher levels of system complexity requiring enhanced Decision Support Tools (DST) to aid humans in the decision making processes. These will rely on accurate predicted aircraft trajectories, provided by advanced Trajectory Predictors (TP). The trajectory prediction process is subject to stochastic effects that introduce uncertainty into the predictions. Regardless of the assumptions that define the aircraft motion model underpinning the TP, deviations between predicted and actual trajectories are unavoidable.
This thesis proposes an innovative method to characterise the uncertainty associated with a trajectory prediction based on the mathematical theory of Polynomial Chaos Expansions (PCE). Assuming univariate PCEs of the trajectory prediction inputs, the method describes how to generate multivariate PCEs of the prediction outputs that quantify their associated uncertainty. Arbitrary PCE (aPCE) was chosen because it allows a higher degree of flexibility to model input uncertainty. The obtained polynomial description can be used in subsequent prediction sensitivity analyses thanks to the relationship between polynomial coefficients and Sobol indices. The Sobol indices enable ranking the input parameters according to their influence on trajectory prediction uncertainty.
The applicability of the aPCE-based uncertainty quantification detailed herein is analysed through a study case. This study case represents a typical aircraft trajectory prediction problem in ATM, in which uncertain parameters regarding aircraft performance, aircraft intent description, weather forecast, and initial conditions are considered simultaneously. Numerical results are compared to those obtained from a Monte Carlo simulation, demonstrating the advantages of the proposed method. The thesis includes two examples of DSTs (Demand and Capacity Balancing tool, and Arrival Manager) to illustrate the potential benefits of exploiting the proposed uncertainty quantification method.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Keywords: Air Traffic Management, Trajectory Based Operations, aircraft Trajectory Prediction, uncertainty quantification, polynomial chaos expansions, aircraft intent uncertainty, weather uncertainty, aircraft performance modelling uncertainty, global sensitivity analysis, Sobol indices.
Subjects: T Technology > TL Motor vehicles. Aeronautics. Astronautics
Colleges/Schools: College of Science and Engineering > School of Engineering > Autonomous Systems and Connectivity
Supervisor's Name: Euan, Mr. McGookin
Date of Award: 2016
Depositing User: Enrique Casado
Unique ID: glathesis:2016-7700
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 24 Oct 2016 08:34
Last Modified: 13 Jul 2017 14:36
URI: https://theses.gla.ac.uk/id/eprint/7700

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year