An Evaluation Index for Cross Ventilation Based on CFD Simulations and Ventilation Prediction Model Using Machine Learning algorithms

Abstract

High-density cities usually surfer deterioration of the urban built environment due to slow or stagnant air movement. This deterioration could be mitigated by improved urban natural ventilation through good urban planning and building design. However, indoor natural ventilation requirements in current building standards only define absolute indoor air speed without considering outdoor conditions. To fill this gap, this paper developed generic urban-scale coupled indoor and outdoor CFD models and proposes a new integrated index (CIOI index) to evaluate indoor natural ventilation potential, which couples both indoor and outdoor wind environments. CIOIv regression models are developed by applying machine learning algorithms based on 3,840 CFD simulation results. The Gradient Boosting (GB) model shows an improved performance in terms of higher correlation (R2=0.8) and lower errors (MAPE=16%) compared with the multivariate linear regression model. Building designers and urban planners can use this CIOI prediction model to quickly check, in the early design stage, the influence of their design variations on indoor ventilation potential.