Content area
Full text
ABSTRACT
The main drawback of Computational Fluid Dynamics (CFD) simulations has been the time and resource consuming nature which is not suitable for real-time applications. In this work, we first generated numerous CFD models of a given indoor space to obtain airspeed, temperature, and mean radiant temperature near an occupant as training data. Several artificial neural networks (ANN) models were trained using this CFD simulated data to approximate near real-time environmental conditions for a given occupant. This trained ANN model approach is a part of a real-time simulation of building operations using a combination of software and real hardware (HVAC equipment) approaches. The preliminary results suggest that the CFD- generated training data and the trained ANN model can accurately approximate such conditions in a real-time application, a method that has great potential in building simulation and building digital twin areas of research.
INTRODUCTION
Indoor airflow is crucial for understanding the thermal condition surrounding the occupants. Computational fluid dynamics (CFD) simulation is the typical tool for indoor airflow distribution investigation.(Chen and Zhai 2004, Sarkar 2008, Ahn et al. 2017) Such investigation is important when one is interested in individual occupant health, comfort, and potential behavior due to these conditions. However, CFD simulations require significant computation resources and time for simulation, and often only representative case studies were simulated due to these constraints. Hence, it is difficult to integrate CFD simulations with decision making tools such as HVAC control and diagnostic applications, as well as realtime simulations which is extremely important in the novel areas of building performance forecasting and building digital twins. Given a the "big-data" natures in these novel research, machine learning (ML) approaches especially the use of artificial neural networks (ANNs) have been prevalent leveraging the vast amount of data produced by building automation systems (BAS) and various Internet of Things (IoT) devices (Mba et al. 2016, Attoue et al. 2018).
For the design and investigation without existing data, however, CFD simulations could provide synthetic data for these ML-based approach. The goal of this investigation is to develop an airflow model for obtaining indoor parameters near occupants using an ANN model trained by CFD simulated data. The trained ANN model would be able to provide approximated these indoor environmental parameters in real-time as...