Analysis of the Performance of Aluminum and Titanium Alloys in Unmanned Aerial Vehicle (UAV) Structures Using Machine Learning Models

Abstract

The dataset used in this research is a well-curated collection of aluminum and titanium alloy samples, compiled from experimentally reported values and reliable literature sources for UAV structural applications. The dataset created using this methodology can simulate engineering practices that use common material systems found in aerospace structures. The dataset provides a comprehensive evaluation of material properties by integrating information from various sources to capture genuine physical, mechanical, fatigue, and operational properties. The study emphasizes the variety of material attributes and the engineering importance of the attributes selected, because these factors determine the accuracy of modeling UAV material selection within a complex multi-criteria system. Each sample's material description includes 12 attributes: density, Young's modulus, yield strength, ultimate tensile strength, fatigue strength, fatigue life, hardness, corrosion resistance, weldability, cost level, and availability. The machine learning models consider both quantitative and qualitative parameters to examine how design requirements interact in nonlinear ways. The selected aluminum and titanium alloys were chosen for this study because they frequently appear in UAV systems, which establishes the dataset as both a theoretical and practical research foundation. The developed machine learning models show improved reliability and generalization, and they provide engineering value for UAV structural design and material selection.