Development of an automatic personal comfort system (APCS) based on real-time thermal sensation prediction

Abstract

Most existing building thermal environment design methods focus on the overall environment without considering individual demands, which can result in inadequate comfort and unnecessary energy consumption. To address individual differences, appropriate measures must be followed at the individual level. This study presents the development of an automatic personal comfort system (APCS) for regulating the local environment of individuals by integrating two personal thermal comfort technologies, personal comfort system (PCS) and personal comfort model (PCM). The equipment enables thermal sensation prediction through infrared skin temperature and environmental parameters measurements, which is used to regulate the corresponding PCS device. The proposed thermal sensation prediction model based on random forest can achieve a prediction accuracy of 69%, while the PCS control strategy comprehensively considers thermal sensation vote (TSV) and airflow acceptability. The APCS performance test experiment involved 25 male adult subjects. The results indicated that PCS operating states were consistent under both manual and automatic control conditions. Moreover, TSV remained within the neutral zone for most participants throughout the experiment. These findings demonstrate that the APCS realized good automatic operation to ensure the comfort of the personnel. The APCS framework holds reference value for research on personal thermal comfort and future thermal environment construction, extending beyond the equipment itself.