ABSTRACT

Building thermal and lighting environment is a key concern in the construction industry. To study the personalized preferences of thermal and lighting environments, the indoor thermal and lighting environment in a typical office building in Ningbo was monitored and analyzed. Building users under different thermal and lighting environment conditions were asked to compare their satisfaction levels, and their environmental personalized preferences were identified and recorded along with the thermal and lighting environment parameters. The statistics show that the comfort zone for users is mostly between 16°C and 28°C. In addition, the usage of indoor lighting systems and the corresponding lighting environment parameters before and after usage were recorded. It can be observed that the average illumination value when using the lighting system is around 60 lux. When the indoor average illuminance reaches 430 lux, the users’ office requirements can be met without using the lighting system. Next, personalized learning of thermal and lighting environment comfort preferences was performed based on the experimental data. By combining Bayesian theory with utility function to define the likelihood function, a personalized learning model for thermal and lighting environment comfort personalized preferences was established. The joint distribution of thermal and lighting comfort environment was analyzed under the condition of thermal and lighting environment coupling. It is found that the observed comfortable lighting range for the occupants is from 300 lux to 6800 lux, and the range of thermal comfort is between corrected PMV (Predicted Mean Vote) values of –0.7 and 1.0.