Genetic algorithms - problem solving in architecture and urbanism

Abstract

To evolutionary-computation researches, the mechanisms of evolution come to be solutions for the most pressing computational problems in many fields. A Genetic algorithm (GA) is a form of evolution that occurs on a computer. These algorithms are appealing sources of inspiration, searching methods that can be used for both - solving problems and modelling evolutionary systems. After describing the basic elements, functionalities and the most important theoretical foundations of GA, this paper will present how GA can be designed to solve different types of problems, including optimization of function, evolving strategies and cellular automata. The accent is interaction between GA and parametric modelling using Grasshopper (the plug-in for the Rhino platform) in the field of architecture and urbanism. This approach finds optimal solution, combination of parameters, which complies given conditions controlled by urbanism's requests. In addition, the influences of the GA control parameters on its solving process are studied. This system has been developed and tasted within the Chair for Mathematics, Architectural Geometry and CAAD, at the University of Belgrade, Faculty of Architecture. This study has partly been completed within the following projects, funded by the Ministry of Education and Science of Serbia: OI174012 and III47014.