A computational framework for synthesising optimum complex structural patterns

作者:Gunaratnam D J; Rosenman M
来源:1st International Conference on Structures and Architecture (ICSA), 2010-07-21 to 2010-07-23.

摘要

The paper describes a computational framework that embeds a number of nature inspired computational paradigms for the synthesis of complex structural patterns to optimise decision criteria such as expressiveness, efficiency, economy and environmental sustainability. The computational paradigms embedded within the framework include cellular automata for generating complex structural patterns of different granularity and complexity using simple rules; a neural network for dynamically learning the preferences of the designer for the expressiveness criterion; and a genetic algorithm for evolving diverse Pareto optimal solutions set. The framework allows the synthesis of both global structural patterns for continuum structures and the more refined patterns for skeletal structures.

  • 出版日期2010