Predictive Modeling and Interaction Analysis of Small Package-Delivery Quadcopter Energy Use

A quadcopter energy model built on the CMU/Rodrigues flight dataset. Quadratic regression beat random forest (R² = 0.911), and an apparent payload × speed interaction turned out to be a scaling artifact rather than real physics.

Constraint

Package-delivery quadcopters need trustworthy energy models from flight data.

Built

Regression models on the CMU/Rodrigues dataset, including quadratic and random forest baselines.

Happened

Quadratic regression reached R² = 0.911; an apparent interaction was a scaling artifact.