Leveraging simulation data to predict thermal comfort using reduce order modelling

Abstract

In today’s world, it’s very important to have an accurate estimation of thermal comfort and indoor air quality at your office spaces, airports, stations, movie theaters, classrooms, etc. For decades, simulation engineers are depending on CFD to calculate the thermal comfort because of growing demands regarding accuracy. It is very time consuming as well as computationally demanding to perform CFD simulations. This limits experimenting various possible designs in HVAC using CFD. From this paper, you will get insights about how machine learning can be implemented to get quicker and accurate thermal comfort prediction before running the actual simulations.