Honors Theses

Date of Award

5-2026

Document Type

Undergraduate Thesis

Degree Name

BS

Department

Mechanical Engineering

Faculty Mentor

Carlos Montalvo, Ph.D.

Advisor(s)

Joseph Richardson, Ph.D. and Mohammadreza Kamaldar, Ph.D.

Abstract

Parafoil–payload systems present significant challenges in modeling and control due to their highly coupled, nonlinear dynamics and sensitivity to atmospheric disturbances. These systems have been thoroughly understood since the mid-1990s, but through the advent of artificial intelligence, new methods for guidance and control remain unexplored. This work implements a high-fidelity 9-DOF simulation framework which captures the multibody interactions between the parafoil, payload, and surrounding airflow, including a stochastic atmospheric model with layered winds and turbulence . Within this environment, LQR and neural network– based controllers are implemented and evaluated under identical conditions. Results show that, while classical controllers perform well in nominal environments, their effectiveness diminishes in dynamic conditions due to modeling limitations. In contrast, the neural network controller demonstrates improved adaptability and robustness, highlighting the potential of data-driven methods for autonomous parafoil guidance and precision landing applications.

Available for download on Wednesday, January 20, 2027

Share

COinS