Integration of psychological parameters into a thermal sensation prediction model for intelligent control of the HVAC systems

Abstract

Conventional thermal comfort models take physiological parameters into account on thermal comfort models. On the other hand, psychological behaviors are also proven as a vital parameter which affects the thermal sensation. In the literature, limited studies which combine both physiological and psychological parameters on the thermal sensation models are exist. To this aim, this study develops a novel Thermal Sensation Prediction Model (TSPM) in order to control the HVAC system by considering both parameters. A data-driven TSPM, which includes Fuzzy Logic (FL) model, is developed and coded using Phyton language by the authors. Two physiological parameters (Mean Radiant Temperature and External Temperature) and one psychological parameter (Emotional Intensity Score (EIS) including Vigour, Depression, Tension with total of 32 subscales) are selected as inputs of the model. Besides the physiological parameters which are decided intentionally considering a manual ventilated building property, the most influencing three sub- psychological parameters on thermal sensation are also selected in the study. While the physiological parameters are measured via environmental data loggers, the psychological parameters are collected simultaneously by the Profile of Mood States questionnaire. A total of 1159 students are participated to the questionnaire at a university study hall between 15th of August 2021 and 15th of September 2022. The results showed that the novel model predicted Thermal Sensation Vote (TSV) with an accuracy of 0.92 of R2. The output of this study may help to develop an integrated Heating Ventilating and Air Conditioning (HVAC) system with Artificial Intelligence - enabled Emulators that also includes psychological parameters.