City Afforestation : Abstracting the Urban Geometries into Tree Structures for Urban Fabric Optimization City Afforestation : Abstracting the Urban Geometries into Tree

Traditional urban design methods, still relying on the static and sectoral approaches, reach their limits to adapt themselves to the growing complexities and dynamics of modern cities. With the flourishing of artificial intelligence, a new approach called "cognitive design computing" has been proposed to utilize both the human creativity and computing power for the development of next generation urban design supporting tools, where Evolutionary Multi-objective Optimization (EMO) is employed for urban synthesis. However, a unified data structure is missing for the genotype representation of different urban features, such as streets, blocks, plots and buildings, so that they can be optimized together. This paper aims to address this problem and is focused on i) the introduction and description of the genotype data structure, and ii) the application of the prototype of the proposed design tool in practice.