Distributed Generative Design Methods
分布式生成设计法
Cellular Automata: A Discrete View for Designing
The built environment plays a vital role in fostering a healthy relationship between humans and nature. Cellular automata (CA) can model natural phenomena in spatio-temporal patterns, enabling the generation of complex spatial layouts based on local morphological rules. This local relationship-driven approach aligns with generative dynamics in natural systems. To tackle the complexities of applying CA rules in ecological urban form, we propose a “multi-layered” CA theoretical model. This model aims to explore architectural designs that prioritize ecological performance, providing a framework for spaces that harmonize with ecological processes and urban dynamics.
细胞自动机:设计的离散视角
建筑环境作为促进人与自然关系健康发展的载体,利用元胞自动机(CA)模拟自然现象及时空模式,为基于局部形态规则生成复杂空间布局提供了可能。其局部关系驱动特性与自然系统中的生成动态相契合。为应对元胞自动机规则在生态城市形态中的复杂性,我们构建了一个“多层次”元胞自动机理论模型,用于探索面向生态性能的建筑形态设计。
Multi-layered CA Models Ι 多层 CA 模型

This project tackles uneven urban vitality by proposing office buildings in interstitial spaces, leveraging decentralization to balance density and enhance activity. Using Cellular Automata and Carcassonne game rules, it introduces a method for automated plan generation with minimal manual input. Constraints added during the design process refine results to meet desired standards, demonstrating self-organizing capabilities. This rule-driven approach promotes design innovation, enabling the creation of unique styles and structures through process manipulation.
本项目通过在城市缝隙空间中建设办公楼,解决城市活力分布不均问题,利用去中心化策略平衡密度并增强活力。结合元胞自动机与《卡卡颂》游戏规则,开发了一种自动生成方案的方法,最大限度减少人工干预。设计过程中逐步添加约束条件,使结果更接近预期标准,展现了自组织方案的生成。这种规则驱动的方法推动设计创新,帮助设计师通过操控过程创造独特风格与结构。
Project Lead: Yiming Liu | 刘燚铭
Publications
The Process of Plan Generation Ι 制定计划的过程

Link