Individual thermal comfort models based on optimized BP neural network algorithms

Thermal comfort plays an important role in human life and it affects occupant satisfaction, health, productivity. Individual differences are not considered traditional control strategies based on temperature setpoints. The reality is that operators often expend more energy to maintain the indoor environment thermal satisfaction of occupancy as well expected. Thus, individual models physiological parameters environmental were presented using back-propagation (BP) neural network. Moreover, we used three training algorithms including Levenberg-Marquardt (L-M), Bayesian Regularization, Scaled Conjugate. We observed L-M algorithm resulted slightly better performance (R=0.96) than other algorithms. precision results suggest BP network effective approach for real-time predicting comfort. In follow-up study, would focus feature engineering (feature selection) introduce appropriate variables (e.g., heart rate) improve model’s accuracy.