Multi-objective early-stage design optimization of multifamily residential projects

Abstract

The aim of this study is to find a building form and plan layout that can be used in the early stages of architectural design, where criteria such as daylight, view, sun-hour, sales area, and cost are optimized according to the different expectations of different housing type users. This study proposes a multi-objective early-stage design optimization for a real estate development project based on the NSGA2 genetic algorithm, considering weighted user preferences for different housing types. The framework is implemented using the platforms Rhino and Grasshopper; Wallacei is used for NSGA2, and Viktor.ai is used to deploy the app. Tested on six sample plots, the model was able to find architecturally optimized results that respond to different user expectations. While the model successfully demonstrated responsiveness to parameters, its focus on Pareto-optimal solutions limited the diversity of unit mixes generated. The model has been tested by professionals on a sample plot and is found to be important for architects and investors to generate ideas at an early stage of architectural design.