Agent-based geometry optimization with Genetic Algorithm (GA) for tall apartment's solar right

In East Asia, whether it is based on cultural or physiological needs, most residents prefer a dwelling space that has a certain amount of natural, direct sunlight to ensure their comfort. For example, in Korea, this need led to the development of building codes requiring most residential buildings to provide sunlight. As a result of following the code, most tall apartment buildings exhibit similar building layouts in which major facades face south, thereby sacrificing design diversity, creativity, and identity. To ease the problem, this paper proposes an alternative method for optimizing a building's access to direct sunlight. The proposed method uses an agent-based geometry control system that sets parameters to control a building in a hierarchical manner, which introduces relatively less control points than the typical parameters used in Genetic Algorithm (GA) optimization. This method allows for both a reduction in the computational cost and time required to determine a solution as well as the possibility to integrate with computational simulation tools that provide the ability of finding a more accurate and objective performance value. Unlike current code, which requires users to follow certain rules as a prerequisite, the proposed method uses performance driven measures to satisfy the requirement thereby allowing for more varied design possibilities to layout tall apartment buildings. (C) 2014 Elsevier Ltd. All rights reserved.