[a]sorted selection : improving building performance and diversity using a new form of interactive evolutionary algorithm

Abstract

There have been numerous studies demonstrating the use of optimization in architecture, yet it has not been adopted in the field, especially in the early stages of design. As buildings become more complex, the use of optimization in design becomes more relevant. Still, there is a gap between the way architects work and the role optimization plays in the design process. This work presents a new methodology, IDEA (interactive diversity-driven evolutionary algorithm), for designers to engage optimization in the early stages of design. IDEA is designed to work in a manner that is comfortable to the designer and easily integrated into the design process. By focusing on similarities and differences of form, designers sort design options into clusters. The clusters are used to develop a diversity-metric that IDEA uses to generate diverse design options that satisfy the objective. This work demonstrates a novel approach to interaction to better assist in early stages design.