Human thermal sensation algorithm modelization via physiological thermoregulatory responses based on dynamic thermal environment tests on males

作者：Li, Weijian; Chen, Jiqing; Lan, Fengchong^*

来源：Computer Methods and Programs in Biomedicine, 2022, 227: 107198.

DOI：10.1016/j.cmpb.2022.107198

• Thermal environment
• Thermal comfort
• Thermal sensation
• Physiological responses
• Modelization

Background and Objectives: Thermal conditions are changeable in cabin space, where occupants could suffer consecutive self-thermoregulation to such changing thermal stresses. Thermal environment man-agement is expected to be purposefully auto-adjustable for the environment by recognizing individual real-time thermal sensations. Current thermal sensation evaluation models are developed for virtual sim-ulations rather than for realistic scenarios, challenging to evaluate human thermal sensation in the field surveys. Methods: The study constructs a human thermal sensation model via human physiological responses to evaluate the human thermal sensation in the actual vehicle environment. The thermal sensation model forms with exponential functions to clarify the relationship between thermal sensation and pulse rate and blood pressure, which successfully expresses the approximately linear trend around neutral sensation and compensates for the end-points bias. The study set up experimental cases to determine the param-eter states in the thermal sensation model. Firstly, subjective thermal sensation scoring was performed by combing with an established seven-point-scale questionnaire survey system for human thermal sen-sation. Wearable sensors are then applied to measure the human physiological response, including blood pressure BP, pulse rate PR and blood oxygen saturation SpO2. Results: The subjects revealed significantly higher pulse rates (positively correlated) and lower blood pressure (negatively correlated) in the warm chamber than in the cool chamber. The defined parame-ter change rate effectively reveals the trend of human thermal sensation and avoids the inconsistency of raw physiological response levels. The change rate in PR and MAP between the thermal sensation in cold-3 and hot +3 is about a 10% difference. Conclusions: Based on the thermal sensation model algorithm, model parameters were fitted by the sub-jects' thermal sensation voting and the change rate of their physiological responses. With the coefficient of determination ( R 2 ) of the regression over 0.8, the proposed thermal sensation model can be employed for human thermal sensation evaluation. The physiological thermoregulatory responses effectively indi-cate the thermal state of the human body and can be used in thermal environments in conjunction with human smart wearable devices.