IMPROVING GENERATIVE GRAMMAR DEVELOPMENT AND APPLICATION THROUGH NETWORK ANALYSIS TECHNIQUES

DS 80-5 Proceedings of the 20th International Conference on Engineering Design (ICED 15) Vol 5: Design Methods and Tools - Part 1, Milan, Italy, 27-30.07.15

Year: 2015
Editor: Christian Weber, Stephan Husung, Gaetano Cascini, Marco Cantamessa, Dorian Marjanovic, Frederico Rotini
Author: K�o�e�n�i�g�s�e�d�e�r�,� �C�o�r�i�n�n�a�; � �S�t�a�n�k�o�v�i�,� �T�i�n�o�; � �S�h�e�a�,� �K�r�i�s�t�i�n�a
Series: ICED
Institution: ETH Zurich, Switzerland
Section: Design Methods and Tools - part 1
Page(s): 167-176
ISBN: 978-1-904670-68-1
ISSN: 2220-4334

Abstract

Design grammars have been successfully applied in numerous engineering disciplines, however, the lack of support for grammar development is seen as one of the major drawbacks of grammatical approaches in Computational Design Synthesis (CDS). In this paper, a method is presented that supports the development and application of design grammars. Concepts from compiler design and transition graphs are used to help designers understand developed grammars in depth. The grammar designer is given feedback on a) the rules, e.g., if there are redundant or do-undo rules, and b) rule application sequences, e.g., which sequences should be preferred or avoided. This feedback can be used to a) improve the grammar, and b) apply it more efficiently. The case study demonstrates how the method is used to analyze a grammar for a sliding tile puzzle. Knowledge learnt on small scale was then successfully applied to solve a larger scale problem. The results show the feasibility of the method and its generality is discussed.

Keywords: Computational Design Synthesis, Conceptual Design, Generative Grammar, Network Analysis, Visualisation

Please sign in to your account

This site uses cookies and other tracking technologies to assist with navigation and your ability to provide feedback, analyse your use of our products and services, assist with our promotional and marketing efforts, and provide content from third parties. Privacy Policy.