UAV conceptual design requires balancing conflicting objectives but still relies on manually selected optimizers that cannot adapt to diverse missions. This project proposes an Automated Algorithm Design framework that automatically composes and configures algorithmic components-mutation, crossover, selection-to build task‑specific optimizers. A meta‑optimizer searches the component space using performance feedback from surrogate models. Expected outcomes include an open‑source prototype, a benchmark suite, and a publication. By automating optimizer design, we aim to shorten design cycles and make high‑performance UAVs more accessible.
