Extract

This research discusses designing an experimental framework to reinterpret the urban form of
self-organizing traditional Islamic cities with computational design tools. These types of ancient urban
tissues are characterized with emergent behaviour that is expressed through the incorporation, in their
spatial structure, of apparent randomness and inherent complex organization. The said framework
highlights some of the important physiological and organisational features of theses urban
environments before reinterpreting to be incorporated in a computational design process. A set of
functional constraints and objectives inspired by characteristics of the aforementioned urban settings
are defined for a form-generation algorithm. This methodology is described through a threefold process
comprising of formulation, generation and evaluation steps that allow for the optimization of the urban
form. While making use of algorithmic design software to define the components for a generative urban
design model, this research also uses multi-objective optimization evolutionary algorithms as a
theoretical basis for the optimization and analysis of this model. An evolutionary computing solver and
Pareto front evaluation methodology are used in order to evaluate and visualize a multitude of
optimized solutions. Built on these features, the development of an experimental urban model means
both to explore the generation of proposed results and to develop the design system behind it.

Keywords: Urban form; multi-objective optimization; Pareto efficiency; generative design.

PDF

Download