Fleet Design Optimization of Package Delivery Unmanned Aerial Vehicles Considering Operations

Authors: Shugo Kaneko and Joaquim R. R. A. Martins

Download paper

Highlight

TODO

Abstract

The conceptual design process of aircraft starts by deciding the representative mission requirements, followed by optimization of design variables to satisfy the given requirements. However, the appropriate mission requirements are not obvious, especially when designing package delivery UAVs (also called drones). The UAVs must accommodate various combinations of package weights and delivery distances. The complexity increases further when designing a heterogeneous fleet of UAVs that serves a large number of customers. This work addresses this problem by solving cou- pled design-operation optimization to find optimal mission requirements and optimal UAV designs simultaneously. We formulate this problem as mixed-integer nonlinear optimization and propose a sequential heuristic algorithm to solve the coupled problem. The benchmark study of the proposed algorithm against a nonconvex branch-and-cut solver shows that the sequential heuristics are effective. We also demonstrate that the simultaneous UAV design and routing optimization reduces the UAV weight across the fleet by more than 12% on average compared to the conventional baseline.