Abstract
The aim of this study is the optimization of user accessibility in architectural design using genetic algorithm. In order to define the problematic, literature is surveyed in several steps: (1) Spatial accessibility focusing user movements, (2) Genetic Algorithm and Optimization in architectural design, and (3) Algorithmic Distance Based Accessibility Model (aDA). Spatial accessibility is the concept which allows user to understand function, organization and spatial relationships and welcome them to participate in activities. In order to understand the concept, methods focusing accessibility in design and planning are studied in detail. The mentioned methods here are space layout planning, space syntax and wayfinding. This paper suggests a different look to these methods in the perspective of spatial accessibility, a further question is asked: Is it possible to optimize the user movement depending on these spatial accessibility criteria? Focusing on user movements, it is aimed to optimize the route of a user to define the distance between two related spaces. Since the architectural space deals with various user types and various functional spaces, optimization is necessary to evaluate the movement. In this respect, genetic algorithm is a useful and well-known method for optimization in architecture. This paper aims to fill the gap in literature to offer a different method as optimization with genetic algorithms in architectural design. The method introduced briefly here is “Algorithmic Distance Based Accessibility Model (aDA)”. It aims to define user and spatial accessibility in architectural design and depending on this, optimization of distances between spaces using genetic algorithm. In this developed model, user and spatial data is scripted depending on a list of criteria and distance optimization is performed by a genetic algorithm. These data are scripted based on a list of criteria that define the scope of the study and evaluated/disevaluated data, as well as tightens the solution pool based on these criteria. The method is evaluated on health campuses, two alternatives are generated and the results are stated.
