An Investigation on Effects of Indoor Temperature and Lighting on Environmental Interaction Perception and Heart Rate Variability Using Machine Learning

People perceive the indoor environment through interacting and integrating various sensory organs. Sixteen volunteers were recruited for an experimental study investigating the cross-effects of temperature and light on environmental perception. The experiment was designed with three independent variables: indoor temperature (24°C, 26°C, and 28 °C), illuminance (100 lx, 300 lx, and 500 lx), and correlated color temperature (CCT) (3000 K, 4000 K, and 6500 K). Subjective questionnaires and heart rate variability (HRV) were used to evaluate volunteers' perceptions of the different indoor thermal-lighting environments. The results show that temperature and CCT have significant effects on thermal sensation. Temperature, illuminance, and CCT have significant second-order effects on thermal sensation. The bright sensation was significantly affected by illumination, and different temperatures affected the sensation of the same illumination. Lighting comfort is significantly affected by CCT, but the temperature has no significant effect on lighting comfort. Environmental perception classification models using SVM and DT algorithms with physical environment parameters and volunteer heart rate variability metrics are established, and all three voting accuracies are above 84%. The established models can be used to provide dynamic regulation strategies for the indoor environment of intelligent buildings.