Evolutionary approach for spatial architecture layout design enhanced by an agent-based topology finding system

TLDR

The study proposes an automatic method to generate spatial architectural layouts from user‑specified architectural programs. The method combines a multi‑agent topology‑finding system that produces topology‑satisfied layouts with an evolutionary optimization process that refines them to meet predefined architectural criteria, thereby narrowing the search space and improving performance. Experiments show that the approach successfully models spatial layouts and supports multi‑floor topologies.

Abstract

This paper presents a method for the automatic generation of a spatial architectural layout from a user-specified architectural program. The proposed approach binds a multi-agent topology finding system and an evolutionary optimization process. The former generates topology satisfied layouts for further optimization, while the latter focuses on refining the layouts to achieve predefined architectural criteria. The topology finding process narrows the search space and increases the performance in subsequent optimization. Results imply that the spatial layout modeling and the multi-floor topology are handled.