Aeromat

Research project at a glance

A sustainable energy future requires that we both do more with less, and that we fully exploit the renewable energy sources we have available. In this project we explore a common thread between these two approaches, developing tools to better explore and understand aerodynamic design. On the one hand our tools can be used to improve the performance of aerodynamic vehicles, and on the other improving our ability to harvest energy from wind. We develop automated methods for the design of complete aerodynamic structures, using machine-learning techniques to guide iterative experimentation with novel designs. We focus on: Optimization of entire structures, rather than iterative improvement on existing designs Human-machine collaborative design exploration, to discover innovative design concepts Inclusion of structural mechanics and fluid structure interaction into the optimization, design, and modeling process Modeling techniques to support these goals, using data-driven approaches to approximate computationally intensive techniques and simulations In particular we face challenges when creating tools which address these issues in tandem, such as: modeling the performance of designs produced with non-traditional parameterizations broad exploration of possible designs in computationally demanding contexts optimization and modeling of aerodynamic and structural properties simultaneously  

Departments and Instituts

Department of Engineering and Communication Department of Computer Science Institute of Technology, Resource and Energy-efficient Engineering (TREE)

Funding type

Publicly funded research

Period

01.10.2016 to 30.09.2019

Project manager at H-BRS

Project Description

A sustainable energy future requires that we both do more with less, and that we fully exploit the renewable energy sources we have available. In this project we explore a common thread between these two approaches, developing tools to better explore and understand aerodynamic design. On the one hand our tools can be used to improve the performance of aerodynamic vehicles, and on the other improving our ability to harvest energy from wind. We develop automated methods for the design of complete aerodynamic structures, using machine-learning techniques to guide iterative experimentation with novel designs.

We focus on:

Optimization of entire structures, rather than iterative improvement on existing designs
Human-machine collaborative design exploration, to discover innovative design concepts
Inclusion of structural mechanics and fluid structure interaction into the optimization, design, and modeling process
Modeling techniques to support these goals, using data-driven approaches to approximate computationally intensive techniques and simulations

In particular we face challenges when creating tools which address these issues in tandem, such as:

modeling the performance of designs produced with non-traditional parameterizations
broad exploration of possible designs in computationally demanding contexts
optimization and modeling of aerodynamic and structural properties simultaneously

Aerodynamic Design Exploration through Surrogate-Assisted
Illumination
(Best Student Paper -- Multidisciplinary Design Optimization)
Gaier, Adam, Alexander Asteroth, and Jean-Baptiste Mouret. AIAA Aviation and Aeronautics Forum 2017

Data-Efficient Exploration, Optimization, and Modeling of Diverse Designs through Surrogate-Assisted
Illumination
(Best Paper -- Complex Systems)
Gaier, Adam, Alexander Asteroth, and Jean-Baptiste Mouret. Genetic and Evolutionary Computation Conference 2017

screen_shot_2017-06-27_at_10.16.27_am.png (DE)

Hierarchical Surrogate Modeling for Illumination Algorithms
Alexander Hagg. Genetic and Evolutionary Computation Conference 2017

Project publications

2021 | 2020 | 2019 | 2018 | 2017

2021

Alexander Hagg: Discovering the preference hypervolume: an interactive model for real world computational co-creativity. Dissertation, Leiden University, 2021
URL BibTeX | RIS

Alexander Hagg: Phenotypic Niching Using Quality Diversity Algorithms. In: Preuss, Epitropakis et al. (Eds.): Metaheuristics for Finding Multiple Solutions, Cham: Springer, 2021, S. 287-315
doi:10.1007/978-3-030-79553-5_12 BibTeX | RIS

2020

Alexander Hagg, Mike Preuss, Alexander Asteroth, Thomas Bäck: An Analysis of Phenotypic Diversity in Multi-solution Optimization. In: Filipič, Minisci et al. (Eds.): Bioinspired Optimization Methods and Their Applications. 9th International Conference, BIOMA 2020, Brussels, Belgium, November 19-20, 2020, Proceedings, 2020, S. 43-55
doi:10.1007/978-3-030-63710-1_4 BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Discovering Representations for Black-box Optimization. In: GECCO '20: Proceedings of the 2020 Genetic and Evolutionary Computation Conference, July 8–12, 2020, Cancún, Mexico, 2020, S. 103-111
doi:10.1145/3377930.3390221 arXiv BibTeX | RIS

Alexander Asteroth, Adam Gaier, Alexander Hagg, Jakob Meng, Andreas Priesnitz, Lea Prochnau, Dirk Reith: AErOmAt Abschlussbericht. Hochschule Bonn-Rhein-Sieg / Institut für Technik, Ressourcenschonung und Energieeffizienz, Hochschule Bonn-Rhein-Sieg, 2020
PDF Download doi:10.18418/opus-4850 urn:nbn:de:hbz:1044-opus-48506 BibTeX | RIS

2019

Alexander Hagg, Martin Zaefferer, Jörg Stork, Adam Gaier: Prediction of neural network performance by phenotypic modeling. In: GECCO '19: Genetic and Evolutionary Computation Conference, Prague, Czech Republic, July 13-17, 2019, 2019, S. 1576-1582
doi:10.1145/3319619.3326815 arXiv BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Are quality diversity algorithms better at generating stepping stones than objective-based search?. In: GECCO '19: Genetic and Evolutionary Computation Conference, Prague, Czech Republic, July 13-17, 2019, 2019, S. 115-116
doi:10.1145/3319619.3321897 BibTeX | RIS

Alexander Hagg, Alexander Asteroth, Thomas Bäck: Modeling User Selection in Quality Diversity. In: GECCO '19: Genetic and Evolutionary Computation Conference, Prague, Czech Republic, July 13-17, 2019, 2019, S. 116-124
doi:10.1145/3321707.3321823 arXiv BibTeX | RIS

2018

Alexander Hagg, Alexander Asteroth, Thomas Bäck: Prototype Discovery Using Quality-Diversity. In: Auger, Fonseca et al. (Eds.): Parallel Problem Solving from Nature – PPSN XV (PPSN 2018). Lecture Notes in Computer Science, vol 11101, 2018, S. 500-511
doi:10.1007/978-3-319-99253-2_40 arXiv BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Data-Efficient Design Exploration through Surrogate-Assisted Illumination. In: Evolutionary Computation, 26(3), 2018, S. 381-410
doi:10.1162/evco_a_00231 PMID arXiv BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Data-efficient Neuroevolution with Kernel-Based Surrogate Models. In: GECCO ’18: Genetic and Evolutionary Computation Conference, July 15–19, 2018, Kyoto, Japan, 2018, S. 85-92 (Best Paper Award)
doi:10.1145/3205455.3205510 arXiv BibTeX | RIS

2017

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Data-Efficient Exploration, Optimization, and Modeling of Diverse Designs through Surrogate-Assisted Illumination. In: GECCO '17: Proceedings of the Genetic and Evolutionary Computation Conference. Berlin, Germany, July 15-19, 2017, 2017, S. 99-106 (Best paper award in the complex systems category)
doi:10.1145/3071178.3071282 arXiv URL BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Aerodynamic Design Exploration through Surrogate-Assisted Illumination. In: 18th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum. 5-9 June 2017, Denver, CO, USA, 2017 (First price in the Multidisciplinary Analysis and Optimization (MDO) Student Paper Competition)
doi:10.2514/6.2017-3330 URL BibTeX | RIS

Alexander Hagg: Hierarchical Surrogate Modeling for Illumination Algorithms. In: GECCO '17: Proceedings of the Genetic and Evolutionary Computation Conference. Berlin, Germany, July 15-19, 2017, 2017, S. 1407-1410
doi:10.1145/3067695.3082495 arXiv BibTeX | RIS

Alexander Hagg, Maximilian Mensing, Alexander Asteroth: Evolving Parsimonious Networks by Mixing Activation Functions. In: GECCO '17: Proceedings of the Genetic and Evolutionary Computation Conference. Berlin, Germany, July 15-19, 2017, 2017, S. 425-432
doi:10.1145/3071178.3071275 arXiv BibTeX | RIS

Adam Gaier, Alexander Asteroth, Jean-Baptiste Mouret: Feature Space Modeling Through Surrogate Illumination. Preprint, 2017 (Published in GECCO '17: Proceedings of the Genetic and Evolutionary Computation Conference, July 2017, Pages 99–106, https://doi.org/10.1145/3071178.3071282)
arXiv doi:10.48550/arXiv.1702.03713 BibTeX | RIS

Results