Multiscale modeling of in-room temperature distribution with human occupancy data: a practical case study

This paper develops a method for modelling of in-room temperature distribution incorporated with data collected by human sensors. This modelling is based on a standard two-dimensional heat diffusion equation with an effective diffusion coefficient. The effective diffusion coefficient is nominally identified from characteristics of air flow inside a room and its architectural design. For modelling multiple time-scale influence of human occupancy on the in-room temperature distribution, two independent parameters—the effective diffusion coefficient and human heat input—of the equation are modulated with the human sensor data that capture spatio-temporal dynamics of the occupancy in high resolution. The developed method is applied to a practical office space in commercial building in Japan so that its effectiveness is demonstrated by comparing numerical simulations of the equation with measured data on temperature.