Design Optimization of Dense Urban Neighbourhood with Computational Simulation and Genetic Algorithm for Improving Outdoor Thermal Environment

Abstract

The outdoor thermal environment is of growing concern in the urban planning and design field, especially in tropical/subtropical regions. The layout and design of urban neighbourhood may have significant environmental impacts. Although computer simulation has been increasingly used to assess outdoor thermal conditions, it is yet to be fully adopted by practitioners in design thinking and processes. This paper proposes a design optimization method to reduce outdoor heat stress and increase the thermal comfortable area in urban spaces. The thermal comfort simulation software CityComfort+ is linked with genetic algorithms to optimize the thermal conditions of open space defined by a cluster of buildings. The method is tested with a hypothetical case in Hong Kong. The algorithm is able to identify high-performing design schemes within a reasonable period, given the site and project constraints. The method can be useful for practice and the further development of simulation-based optimization methods.