Shape Generation Using Pareto Genetic Algorithms: Integrating Conflicting Design Objectives in Low-Energy Architecture

Abstract

The Generative Design System [GDS] presented in this paper was developed to assist designers in researching low-energy architecture solutions. The GDS has the capability to evolve architectural forms that are energy-efficient, while complying to design intentions expressed by the architect and responding to conflicting objectives. To achieve this evolutionary development, the system integrates a search and optimization method [Genetic Algorithm], building energy simulation software [DOE2.1E], and Pareto multicriteria optimization techniques. The GDS adaptively generates populations of alternative solutions, from an initial schematic layout and a set of rules and constraints designed by the architect to encode design intentions. The two conflicting objective functions considered in this paper are maximizing daylighting use and minimizing energy consumption for conditioning the building. The GDS generated an uniformly sampled, continuous Pareto front, from which six points were visualized in terms of the proposed architectural solutions.