A genetic algorithm approach to planning of buildings in urban green space

Ziyu Tong,Wowo Ding

DOI: https://doi.org/10.1109/ICETCE.2011.5775697

2011-01-01

Abstract:A certain buildings are required in urban green space, and their area depends on the type of park and area of land. However, the regulations don't specify the distribution pattern of buildings in green space. When the green space is very large, the area of buildings would reach a large scale. Such large-scale building or group will impact the local ecological environment and landscape. Therefore, the body sizes and distribution patterns of buildings need be treated carefully. In the previous research, we had established a quantitative index - the Index of Distribution (IOD) - to describe the distribution pattern of buildings in greenspace. The index could be used to assess the quality of plan of buildings accurately. However, it's hard to be used to conduct the planning directly. Based on the genetic algorithm, the paper presents a new model which can automatically generate the plan of buildings that corresponds with the specific IOD. In the model, the IOD is adopted as the fitness function; the coordinates of locations of all buildings are coded as the gene of the chromosome. Compared with normal genetic algorithm, the model constructs the chromosome with real number instead of binary data. With such feature, the model can be more effective to respond to the flexibility of location of building. The result of the model can be used as the valuable reference to the planning of buildings in green space. © 2011 IEEE.