A New Shape Generation Framework Based on Machine Learning and Topology Optimization

512

Chapter Title

A New Shape Generation Framework Based on Machine Learning and Topology Optimization

Book Title

Proceedings of International Conference on Engineering, Science and Technology 2021

Chapter Author(s)

Nikos Ath. Kallioras, Nikos D. Lagaros

Editors

Dr. Mack Shelley, Dr. Valarie Akerson

ISBN

978-1-952092-24-4

Pages

48-59

Abstract

Lately, an algorithmic tool, known as Generative Design, that supports products’ design has been introduced in several industries with manufacturing procedures. Generative Design can be described as the technology that focuses on the generation of plethora of designs that all respect designer-set criteria such as loading conditions, support conditions, etc. Due to this feature, generative design can be used as an intuition creating tool that actually suggests several rough prototypes to the designer who can use them as inspiration for the final prototype. In this work, a novel shape generation framework based on topology optimization, machine learning and image editing is proposed, aiming at performing generative design in architectural design. In detail, the proposed framework constitutes a combination of Solid Isotropic Material with Penalization (SIMP) (Bendsøe, 1989), Long Short-Term Memory networks (LSTM) (Hochreiter & Schmidhuber, 1997) and various image filters. SIMP is used for optimizing the shape according to the designer’s criteria and constraints while LSTMs and image filtering are used for shape differentiation and process acceleration. The proposed framework is tested over a number of topology optimization problems used as benchmark tests in modern literature.


Citation

Kallioras, N. A. & Lagaros, N. D. (2021). A new shape generation framework based on machine learning and topology optimization. In M. Shelley & V. Akerson (Eds.), Proceedings of IConEST 2021-- International Conference on Engineering, Science and Technology (pp. 48-59), Chicago, USA. ISTES Organization.



Other Chapters


Cover Page & Contents

More Info Download


Virtual Teaching/Learning on Engineering Graphics Course in COVID-19 Pandemic

Xinchuan Liu, Peng Cheng

More Info DownloadPages: 1-11


Robust Multi-Objective Optimal Design of a Racing Car Suspension System

Muhammad Ali Khan, Yousef Sardahi, Carlos Ignacio Hernández Castellanos

More Info DownloadPages: 12-25


Neural Network Modeling and Optimizing of the Agglomeration Process

Gulnara Abitova, Leila Rzayeva, Tansuly Zadenova

More Info DownloadPages: 26-35


Design of an Integrated Data Acquisition System for Aero Engine Testing Using LabVIEW® Virtual Instrumentation

V Prabakar, K Srinivasan

More Info DownloadPages: 36-47


A New Shape Generation Framework Based on Machine Learning and Topology Optimization

Nikos Ath. Kallioras, Nikos D. Lagaros

More Info DownloadPages: 48-59


Minors Would Be Safe Browsing of Internet through Intelligent Systems

Ana Elena Taboada Montes de Oca, Richard de Jesús Gil Herrera

More Info DownloadPages: 60-69


Work-Based Experiences for School Pupils in Universities

Pierfrancesco Riccardi

More Info DownloadPages: 70-77


Event & News
Events linked with ISTES

Conferences linked with International Society for Technology, Education and Science (ISTES): International Conference on Social and Education Sciences (IConSES) – USA: www.iconses.net Int...

January 19, 2022

View details »

Publications linked with ISTES

Journals linked with International Society for Technology, Education and Science (ISTES): International Journal of Education in Mathematics, Science and Technology (IJEMST): www.ijemst.net ...

November 26, 2020

View details »

Book Chapter Publication & Call for Book Chapters on Artificial Intelligence (AI) in Education

Call for Book Chapters on Artificial Intelligence (AI) in Education   The International Society for Technology, Education and Science (ISTES)  is interested in chapters with a focus on...

January 03, 2020

View details »

Social Media