A dynamic two-partition model of sleep thermal comfort in hot–humid climates

The indoor thermal environment significantly influences sleep quality, particularly under hot and humid climate conditions. Existing models for predicting the sleep thermal environment typically assume a constant metabolic rate and employ oversimplified heat exchange mechanisms, thereby failing to capture the dynamic characteristics of sleep physiology and the complexities of human heat transfer. This study developed a novel thermal comfort model for sleep, the dynamic predicted mean vote-two partition (DPMV-2P), which integrates a dynamic metabolic rate function, M ( t ), with a two-partition heat exchange framework. Validation results demonstrate that the DPMV-2P model exhibits superior predictive accuracy, achieving R 2 up to 0.999 and reducing MAE by up to 85% compared with the predicted mean vote (PMV) and sleep predicted mean vote (SPMV) models, and EnergyPlus simulations under a variable air volume (VAV) system showed that DPMV-2P predictions remained stable within 0.4 to 0.6, closest to the thermal comfort zone (−0.5 to 0.5), compared with SPMV predictions of 0.5 to 1.0. The model provides a reproducible and practical framework for assessing sleep thermal comfort in hot and humid climate conditions.