Multi-objective diagrid facąde optimization using differential evolution

Abstract

Facądes constitute one of the fundamental systems of contemporary buildings. They serve multiple purposes such as to ensure proper indoor climate to provide sufficient daylight but also to create a desirable architectural image. Integration of these aspects makes facąde design a complex task that requires significant effort in order to achieve well-performing results. It is thus desirable that systematic approaches to facąde design are developed. In this study we consider facąde design as a multiobjective optimization problem integrating diverse design criteria namely indoor daylight distribution structural performance and cost. We evaluate design performance by making use of simulation. Consequently we use Differential Evolution (DE) to search for best-tradeoff solutions. We compare the performance of two DE variants using the Hypervolume metric and also through qualitative inspection. We report facąde designs that demonstrate interesting and often unexpected features concluding that the proposed approach may lead to a novel more integrated design process. © 2017 Elsevier B.V. All rights reserved.