Student Name: Anushka S. Moharir

Advisor: Dr. Dimitri Mavris

Milestone: PhD Thesis Proposal

Degree Program: Aerospace Engineering

Title: Adaptive Aircraft Trajectory Optimization Under Environmental and Operational Uncertainty

Abstract: Aircraft arrival operations in high-density terminal airspace are subject to significant environmental and operational uncertainty, including wind variability, traffic sequencing constraints, and air traffic control interventions. While modern descent procedures are optimized to reduce fuel burn and community noise exposure, they are typically designed under nominal conditions and implemented in an open-loop manner. In practice, deviations from assumed conditions degrade performance and reduce robustness. These limitations motivate the need for adaptive trajectory optimization frameworks capable of responding dynamically to evolving uncertainty. Reinforcement learning has emerged as a promising approach for adaptive trajectory generation due to its closed-loop decision structure. However, in safety-critical aerospace systems, RL-based methods face structural challenges related to robustness guarantees, interpretability, and constraint enforcement within physics-based simulation environments. Deterministic optimal control methods provide physical consistency but lack adaptivity, while surrogate modeling improves computational efficiency yet is often weakly integrated into adaptive control architectures. A unified framework that combines robustness, interpretability, and closed-loop adaptivity remains lacking. This thesis proposes a physics-grounded modeling and simulation framework for adaptive aircraft trajectory optimization under environmental and operational uncertainty. The research investigates three structural hypotheses: (1) uncertainty-aware deterministic formulations improve robustness without significant nominal performance degradation; (2) physics-informed surrogate models enable efficient adaptive re-computation under dynamic constraints; and (3) reinforcement learning with physics-grounded reward structures and explicit constraint enforcement achieves interpretable and stable closed-loop trajectory control. A reduced-order longitudinal flight dynamics model and controlled uncertainty scenarios are used to evaluate performance stability under perturbation and adaptive effectiveness. By reframing arrival trajectory optimization as a structured, uncertainty-aware feedback problem grounded in physical modeling, this work advances scalable methodologies for robust adaptive control in next-generation aviation systems.

Date and time: 2026-03-27, 1 pm - 4 pm

Location: Collaborative Design Environment (CoVE), Weber (SST II)

Committee:
Dr. Dimitri Mavris (advisor), School of Aerospace Engineering
Prof. Dimitri Mavris, School of Aerospace Engineering
Prof. Daniel Schrage, School of Aerospace Engineering
Prof. Kai James, School of Aerospace Engineering
Dr. Ameya Behere, School of Aerospace Engineering
Dr. Angela Campbell, FAA