Design Automation and Optimization Methodology for Electric Multicopter UAVs

TL;DR

This paper introduces an automated, two-stage design methodology for electric multicopter UAVs that optimizes components and parameters efficiently based on specific performance requirements.

Contribution

It presents a novel offline-online algorithm framework that automates multicopter design, improving speed and accuracy over traditional iterative methods.

Findings

The method achieves high precision in design optimization.

It significantly reduces computation time compared to brute-force search.

Experiments validate the effectiveness and practicality of the approach.

Abstract

The traditional multicopter design method usually requires a long iterative process to find the optimal design based on given performance requirements. The method is uneconomical and inefficient. In this paper, a practical method is proposed to automatically calculate the optimal multicopter design according to the given design requirements including flight time, altitude, payload capacity, and maneuverability. The proposed method contains two algorithms: an offline algorithm and an online algorithm. The offline algorithm finds the optimal components (propeller and electronic speed controller) for each motor to establish its component combination, and subsequently, these component combinations and their key performance parameters are stored in a combination database. The online algorithm obtains the multicopter design results that satisfy the given requirements by searching through the component combinations in the database and calculating the optimal parameters for the battery and airframe. Subsequently, these requirement-satisfied multicopter design results are obtained and sorted according to an objective function that contains evaluation indexes including size, weight, performance, and practicability. The proposed method has the advantages of high precision and quick calculating speed because parameter calibrations and time-consuming calculations are completed offline. Experiments are performed to validate the effectiveness and practicality of the proposed method. Comparisons with the brutal search method and other design methods demonstrate the efficiency of the proposed method.