The University of Southampton
University of Southampton Institutional Repository

# The Development of a Conceptual Design Algorithm of Multicopters and the Study of its Applications

Guo, Yangzi (2020) The Development of a Conceptual Design Algorithm of Multicopters and the Study of its Applications. University of Southampton, Masters Thesis, 257pp.

Record type: Thesis (Masters)

## Abstract

Multicopters is a large branch of unmanned aerial vehicles (UAVs). Due to the high agility and simple structure of multicopters, they are popularly used in many commercial and civilian applications. However, for a long time, there have existed few systematic algorithms to design multicopters. Most manufacturers and amateurs tend to design multicopters based on experience
rather than engineering principles.

In this thesis, a conceptual design algorithm of multicopters is developed which has a better confidence of estimation than in previous works. The main part of this algorithm is a process which determines the minimum gross take-off mass (퐺푇푂푀) of a certain combination of design parameters. This process calls the layer_2 function; it begins with a guessed 퐺푇푂푀, considers
the given mission requirements, and then calculates the required power for different flight conditions. Next, the required performances of each component are calculated based on the power and are used for estimating the components’ mass and price. At the end, the total mass is summed as the estimated 퐺푇푂푀. If the error between the initial 퐺푇푂푀 and the estimated 퐺푇푂푀
is close to zero, the initial 퐺푇푂푀 can be considered the minimum valid 퐺푇푂푀. Outside of this process, an external loop proceeds through all the design parameter combinations and develops a pool of valid designs (can converge to the minimum 퐺푇푂푀). Optimised designs can be filtered out with given targets (e.g. 푀푇푂푀, total price, endurance and the maximum speed, among others).

Compared with the previous methods of multicopter conceptual design, this algorithm offers the
following innovations:
• This algorithm considers the entire aircraft rather than focussing on a single separated system.
• The forest-tree-bagging technique is applied on the regression process of components, achieving better accuracy of component selection.
• The methods used in previous works are normally valid for only a narrow range of products, while this algorithm is valid for small (approximately 100 g) to relatively large (approximately 30 kg) multicopters.
• The algorithm can optimise not only the maximum take-off mass (푀푇푂푀) but also the price, endurance, total energy consumed and maximum speed.

This algorithm is validated for four aspects: the components estimation, total mass estimation, function output and optimisation process. Sensitivity analyses are then conducted for some selected parameters.

With the aforementioned advantages, this validated algorithm can be used as a tool for studying some specific flight mission problems. The algorithm can not only determine the most suitable design for a specific mission but can also be used to evaluate the extent to which a design is matched to a mission, compare several designs for a mission or even help to improve the mission. This is one of the primary contributions of this thesis, since such a tool for multicopters has not yet been seen.

An example scenario is provided at the end of this thesis. For a long-range mission involving several multicopters, the signal should remain connected throughout the mission. One multicopter would serve as the leader to complete the mission, while the others will remain at each waypoint to function as repeaters. For such a mission, the multicopters can be organised into different configurations, similar to how a mother carries several children. Three configurations are proposed and compared based on several aspects.

Text
Final thesis - Version of Record
Text
Permission to deposit thesis - form - Frank Guo
Restricted to Repository staff only

Published date: 2020

## Identifiers

Local EPrints ID: 453028
URI: http://eprints.soton.ac.uk/id/eprint/453028
PURE UUID: 1070c05f-5ba9-4000-8d99-603756578d12
ORCID for Yangzi Guo: orcid.org/0000-0002-6207-4975
ORCID for David Toal: orcid.org/0000-0002-2203-0302

## Catalogue record

Date deposited: 07 Jan 2022 17:41

## Contributors

Author: Yangzi Guo