A reduced order model for turbulent flows in the urban environment using machine learning. (15th January 2019)

• Record Type: Journal Article
• Authors: Xiao, D.; Heaney, C.E.; Mottet, L.; Fang, F.; Lin, W.; Navon, I.M.; Guo, Y.; Matar, O.K.; Robins, A.G.; Pain, C.C.
• Is Part Of: Building and environment. Volume 148(2019)
• Page Start: 323
• Page End: 337
• Publication Date: 2019-01-15
• Subjects: Non-intrusive reduced order modelling -- Urban flows -- Proper orthogonal decomposition -- Machine learning -- Gaussian process regression -- Operational modelling; Buildings -- Environmental engineering -- Periodicals; Building -- Research -- Periodicals; Constructions -- Technique de l'environnement -- Périodiques; Electronic journals; 696
• DOI: 10.1016/j.buildenv.2018.10.035
• Languages: English
• ISSNs: 0360-1323

Abstract

Abstract: To help create a comfortable and healthy indoor and outdoor environment in which to live, there is a need to understand turbulent air flows within the urban environment. To this end, building on a previously reported method [1], we develop a fast-running Non-Intrusive Reduced Order Model (NIROM) for predicting the turbulent air flows found within an urban environment. To resolve larger scale turbulent fluctuations, we employ a Large Eddy Simulation (LES) model and solve the resulting computational model on unstructured meshes. The objective is to construct a rapid-running NIROM from these results that will have 'similar' dynamics to the original LES model. Based on Proper Orthogonal Decomposition (POD) and machine learning techniques, this Reduced Order Model (ROM) is six orders of magnitude faster than the high-fidelity LES model and we demonstrate how 'similar' it can be to the high-fidelity model by comparing statistical quantities such as the mean flows, Reynolds stresses and probability densities of the velocities. We also include validation of the high-fidelity model against data from wind tunnel experiments. This paper represents a key step towards the use of reduced order modelling for operational purposes with the tantalising possibility of it being used in place of Gaussian plume models, and the potential for greatly improved model fidelity and confidence. Highlights: A Non-Intrusive Reduced Order Model (NIROM) is constructed using a Gaussian Regression Process machine learning method. NIROM is used to predict the turbulent flows found within an urban neighbourhood of the Elephant and Castle area of London. First implementation of such a model into an advanced 3D unstructured finite element mesh fluid model (Fluidity). The model is several orders of magnitude faster than the high-fidelity Large Eddy Simulation (LES) model.