Wenxin Zhang
(Advisor: Prof. Dimitri N. Mavris]

will propose a doctoral thesis entitled,

A Data-driven Methodology for Real-time Aircraft Trajectory Analysis to Reduce Mid-air Collision Risk in Terminal Airspace

On

Wednesday, March 15 at 8:00 a.m.
Weber Space Science and Technology Building 304

Click here to join the meeting

Abstract
A mid-air collision refers to an aviation accident that occurs when two aircraft make contact with each other while in flight. It is considered to be one of the most catastrophic aviation accidents in history and continues to pose a significant safety concern in present-day operations. Currently, Air Traffic Control (ATC) is the primary source to ensure safe separation between aircraft and prevent mid-air collisions. It relies mostly on the expertise of human operators, commonly referred to as Air Traffic Controllers (ATCOs), to carry out mission-critical tasks with heavy workloads. The projected growth of aviation in terms of both traffic volume and diversity, particularly in terminal airspace, presents significant challenges to the existing ATC system, as the burden upon ATCOs may surpass their capacity, thereby compromising their performance. To adapt to future aviation demands and uphold a superior level of safety, ATC is progressively introducing automated systems that help ATCOs make the shift from manual to supervisory tasks. This dissertation is motivated by the necessity for sophisticated analysis and automated decision support regarding air traffic in terminal airspace.

Traditional physics-based dynamical models are limited in their ability to handle the vast scale and intricacy of anticipated air traffic demands. However, data-driven techniques such as machine learning have demonstrated superior performance with the availability of sufficient data. With the utilization of Global Navigation Satellite System (GNSS) technologies, particularly Automatic Dependent Surveillance–Broadcast (ADS-B), real-time and extensive historical operational data are accessible to ATC. Consequently, this study proposes a novel data-driven methodology to conduct real-time analysis of aircraft trajectories, aiming to mitigate the risk of mid-air collisions in terminal airspace.

The proposed methodology comprises three main steps: (1) traffic flow identification and recognition, (2) trajectory prediction, and (3) conflict detection. In the first step, historical trajectory data will be clustered to identify traffic flows in terminal airspace, and then classification models will be developed to recognize the traffic flows of specific flights in real-time. The second step involves leveraging deep-learning techniques to train models for trajectory prediction. Finally, spatial data structures will be employed to enable scalable airspace-level conflict detection based on predicted aircraft trajectories.

Committee

• Prof. Dimitri Mavris – School of Aerospace Engineering (advisor)
• Prof. Daniel P. Schrage – School of Aerospace Engineering
• Dr. Tejas Puranik – NASA Ames Research Center (USRA)
• Dr. Alexia Payan – Research Engineer, Aerospace Systems Design Laboratory