Comfort band for adaptive model based on quadratic regression and probit analysis

¹ Graduate School of Environmental and Information Studies, Tokyo City University, Japan.
² Tokyu Fudosan Holdings Co., Tokyo, Japan.

^* Corresponding author: g2293103@tcu.ac.jp

Abstract

Many researchers focus on developing adaptive models without concerning the comfort band as most of them only refer to the ASHRAE or other global standards. Therefore, a field measurement was conducted at a residential building in Tokyo, Japan for 2 years in order to develop the comfort band by using quadratic regression and probit analysis for the adaptive thermal comfort model. A total of 32,988 thermal sensation votes (TSV) from thermal comfort surveys were collected. The results showed that the mean thermal sensation vote was 4.0 in free-running (FR), 3.7 in heating (HT), and 4.2 in cooling (CL) modes which indicates that the residents were generally satisfied with the condition of indoor environment in the dwellings. This may be because the residents are well-adapted to the local climate and culture. The comfort temperature during FR, HT, and CL modes were 23.7ºC, 20.9ºC, and 26.8ºC, respectively. The mean and standard deviation of the difference between indoor and comfort temperature (ΔT) is 0.2±1.2°C. In developing the comfort band for the adaptive model in this study, quadratic regression and probit analysis utilized the data of ΔT. The appropriate comfort band was fixed as ±1.5°C and ±2.0°C for 90% and 80% limits of the comfort band. These results are crucial when creating reliable standards and guidelines for building design or indoor environmental quality assessment.