School of Engineering Department of Mechanical and Aerospace Engineering 188 Ambular: An Emergency Response Drone Supervisor: Larry LI / MAE Student: BHULLAR Jobanpreet Singh / AE Course: UROP 1100, Fall UROP 2100, Spring UROP 3200, Summer This study presents a novel framework for minimizing the drag coefficient Cd of the Ambular electric vertical takeoff and landing (eVTOL) vehicle, while maximizing the lift coefficient Cl. Leveraging wind tunnel experiments and XFOIL software for aerodynamic analysis and Deep Reinforcement Learning (DRL) for solution exploration, we develop an optimization pipeline that iteratively evaluates different airfoil geometries under different flow and geometry constraints. The DRL agent, trained through a reward function penalizing drag and lift deviations, navigates the complex design space to identify promising designs. Results demonstrate a significant reduction in Cd (up to 8.7% at Re=700,000 and AOA=2°) with negligible compromise in lift performance. This approach not only enhances the aerodynamic efficiency and energy savings of the Ambular platform but also establishes a robust, automated methodology for constrained aerodynamic shape optimization applicable to a broad range of engineering applications.
RkJQdWJsaXNoZXIy NDk5Njg=