Interactive optimisation for high-lift design.
| dc.contributor.advisor | Savill, Mark A. | |
| dc.contributor.advisor | Kipouros, Timoleon | |
| dc.contributor.advisor | Savvaris, Al | |
| dc.contributor.author | Balasani, Myrtle Julia | |
| dc.date.accessioned | 2023-05-23T14:24:01Z | |
| dc.date.available | 2023-05-23T14:24:01Z | |
| dc.date.issued | 2018-10 | |
| dc.description.abstract | Interactivity always involves two entities; one of them by default is a human user. The specialised subject of human factors is introduced in the context of computational aerodynamics and optimisation, specifically a high-lift aerofoil. The trial and error nature of a design process hinges on designer’s knowledge, skill and intuition. A basic, important assumption of a man-machine system is that in solving a problem, there are some steps in which the computer has an advantageous edge while in other steps a human has dominance. Computational technologies are now an indispensable part of aerospace technology; algorithms involving significant user interaction, either during the process of generating solutions or as a component of post-optimisation evaluation where human decision making is involved are increasingly becoming popular, multi-objective particle swarm is one such optimiser. Several design optimisation problems in engineering are by nature multi-objective; the interest of a designer lies in simultaneous optimisation against two or more objectives which are usually in conflict. Interactive optimisation allows the designer to understand trade-offs between various objectives, and is generally used as a tool for decision making. The solution to a multi-objective problem, one where betterment in one objective occurs over the deterioration of at least one other objective is called a Pareto set. There are multiple solutions to a problem and multiple betterment ideas to an already existing design. The final responsibility of identifying an optimal solution or idea rests on the design engineers and decision making is done based on quantitative metrics, displayed as numbers or graphs. However, visualisation, ergonomics and human factors influence and impact this decision making process. A visual, graphical depiction of the Pareto front is oftentimes used as a design aid tool for purposes of decision making with chances of errors and fallacies fundamentally existing in engineering design. An effective visualisation tool benefits complex engineering analyses by providing the decision-maker with a good imagery of the most important information. Two high-lift aerofoil data-sets have been used as test-case examples; a multi-element solver, an optimiser based on swarm intelligence technique, and visual techniques which include parallel co-ordinates, heat map, scatter plot, self-organising map and radial coordinate visualisation comprise the module. Factors that affect optima and various evaluation criteria have been studied in light of the human user. This research enquires into interactive optimisation by adapting three interactive approaches: information trade-off, reference point and classification, and investigates selected visualisation techniques which act as chief aids in the context of high-lift design trade studies. Human-in-the-loop engineering, man-machine interaction & interface along with influencing factors, reliability, validation and verification in the presence of design uncertainty are considered. The research structure, choice of optimiser and visual aids adapted in this work are influenced by and streamlined to fit with the parallel on-going development work on Airbus’ Python based tool. Results, analysis, together with literature survey are presented in this report. The words human, user, engineer, aerodynamicist, designer, analyst and decision-maker/ DM are synonymous, and are used interchangeably in this research. In a virtual engineering setting, for an efficient interactive optimisation task, a suitable visualisation tool is a crucial prerequisite. Various optimisation design tools & methods are most useful when combined with a human engineer's insight is the underlying premise of this work; questions such as why, what, how might help aid aeronautical technical innovation. | en_UK |
| dc.description.coursename | PhD in Aerospace | en_UK |
| dc.identifier.uri | https://dspace.lib.cranfield.ac.uk/handle/1826/19704 | |
| dc.language.iso | en | en_UK |
| dc.rights | © Cranfield University, 2015. All rights reserved. No part of this publication may be reproduced without the written permission of the copyright holder. | |
| dc.subject | Interactive optimisation | en_UK |
| dc.subject | multi-objective | en_UK |
| dc.subject | high-lift design | en_UK |
| dc.subject | human-in-the-loop | en_UK |
| dc.subject | human factors | en_UK |
| dc.subject | decision-maker | en_UK |
| dc.subject | digital human | en_UK |
| dc.subject | particle swarm | en_UK |
| dc.subject | modelling and stimulation | en_UK |
| dc.subject | transonic wing design | en_UK |
| dc.subject | visual analytics | en_UK |
| dc.subject | interactive visualisation | en_UK |
| dc.subject | virtual product engineering | en_UK |
| dc.title | Interactive optimisation for high-lift design. | en_UK |
| dc.type | Thesis | en_UK |