COMPUTATIONAL DESIGN SUPPORT : PERFORMANCE OPTIMIZATION VIA AUTOMATIC GENERATIVE ALGORITHM

Abstract

This dissertation aims to develop a design support toolthat integrates simulation within design processes, to aid architects in optimizing design options by considering lighting and thermal performance. A gap analysis between digital design modeling tools and simulation tools is conducted to better understand the challenges of integration and to map a development framework for a new design support software tool. In this tool, Industry Foundation Classes1 (IFC) via Architecture Revit2 is chosen to achieve Building Information Modeling (BIM) and software interoperability, while EnergyPlus3 is used for performance simulation and analysis. A user interface (UI) is proposed and prototyped in order to significantly reduce time, effort and cost associated with conducting simulation. This dissertation shows how the technology in the prototype software makes generative optimization possible, by parametrically generating design options, automatically evaluating the performance of each option, and identifying the most viable options using optimization techniques.