| Title |
Development of the Estimating Process for the real-time Clothing Insulation of Multiple Occupants |
| Authors |
최은지(Choi, Eun Ji) ; 윤지영(Yun, Ji Young) ; 김남현(Kim, Nam Hyeon) ; 문진우(Moon, Jin Woo) |
| DOI |
https://doi.org/10.5659/JAIK.2023.39.2.257 |
| Keywords |
Thermal comfort; Occupant-centric control; Predicted mean vote; Clothing insulation |
| Abstract |
Information on clothing insulation (CLO) worn by individual occupants is essential for comfort-based control in buildings. It is necessary to
develop a method for estimating each occupant's real-time actual clothing information in situations with multiple occupants. This study aims
to suggest a novel approach to estimate the CLO of multiple occupants in real-time and confirm its practical application with the help of a
performance evaluation in a test-bed. A process that combined the person detection model and the CLO estimation model was proposed.
Experiments to estimate CLO worn by multiple persons in a test-bed were conducted to train the model and evaluate its performance at each
stage of the process. As a result, each model of the process demonstrated an average person detection accuracy of 95% and a CLO
estimation accuracy of 91% for six different clothing combinations. Consequently, these experimental findings confirmed the feasibility of
estimating a person's clothing insulation in an indoor environment where multiple people exist. |