Development of an integrated approach for the inverse design of built environment by a fast fluid dynamics-based generic algorithm

• Published in: Building and environment Vol. 160; p. 106205
• Main Authors: Xue, Yu, Liu, Wei, Wang, Qian, Bu, Fang
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.08.2019; Elsevier BV
• Subjects: Aerodynamics; Air quality; Algorithms; Built environment; Computational fluid dynamics; Computer applications; Computer simulation; Crossover rate; Energy consumption; Fluid dynamics; HVAC equipment; Hydrodynamics; Inverse design; Mutation; Mutation rate; Mutation rates; Optimal design; Population number; Population size; Thermal comfort; Thermal environments; Urban environments; Ventilation; Population size; Mutation rate; Crossover rate; Optimal design
• ISSN: 0360-1323; 1873-684X; 1873-684X
• DOI: 10.1016/j.buildenv.2019.106205

It is essential to further design built environments with improved thermal comfort level, air quality, and reduced energy consumption of the HVAC system. The CFD-based GA was able to identify the global optimal design, but this method requires numbers of CFD simulations which is time consuming. Besides, there is no general rule in determining the critical parameters of GA, such as population size, mutation rate, and crossover rate. Therefore, this study adopted the FFD instead of CFD and developed the FFD-based GA in OpenFOAM. By testing the FFD-based GA in designing the thermal environment in an office with displacement ventilation, it was found that the FFD-based GA had the similar performance with that of the CFD-based GA and saved more than 75% computational effect. Making use of the efficiency of the FFD-based GA, this investigation tested the effect of population size, mutation rate, and crossover rate on the inverse design by GA. In the same design case, the appropriate population size was n=16 and mutation rate was m=0.1, while the crossover rate had no general effect on the inverse design. •A fast fluid dynamics (FFD)-based genetic algorithm (GA) was developed using OpenFOAM.•Compared with CFD-based GA, FFD-based GA saved more than 75% computation effort with similar design accuracy maintained.•The effect of population size, mutation rate, and crossover rate on inverse design was tested.