Generative design optimization in urban planning – walkability-optimized city concept
This paper presents an urban analysis and design workflow using Rhinoceros/Grasshopper with evolutionary solver Galapagos and add-ons: Shortest Walk (ShortWalk) and Elk. The research is based on the authentic urban situation in the city of Irkutsk and was conducted within the 15th session of the International Baikal Winter University of Urban Planning 2014. The aim of the research is to propose a workflow, which reforms some part of the existing urban structure into walkable neighborhood. The study uses a walkability calculator to estimate the walk scores for the residential buildings within the design perimeter and applies genetic algorithms to point out the location of the new amenities to maximize the average of the particular walk scores. As a result, the conceptual project of walkable-optimized neighborhood is proposed. The conclusions show the limitations and potentials of the applied algorithm, discuss the resulting project and outline future research.
Key words: generative design, optimization, walkability, walk score, genetic algorithms
full article - PDF file