A Data‐Driven Decision Framework for Optimal PECS Selection to Meet Heterogeneous Cooling Demands in Dynamic Indoor Environments: A Case Study in a University Classroom

With recent advances in IT, it has become possible to individually control the indoor environment for each room in buildings. However, there are differences in thermal preferences and local heat distribution between occupants and between. To settle such problems, personalized environmental control systems (PECS) have emerged these days as a solution, which provides the advantages of portability, easy installation, and operation as local environmental control devices. However, research on the optimal selection of cooling PECS devices still remains limited. Therefore, this study explored to develop a decision‐making framework to select optimal PECS—among typical PECS devices (i.e., desk fan, body fan, cooling cushion, and neck band)—based on individual characteristics and activities. To avoid large data requirements of conventional machine learning, the proposed framework uses association rule mining (ARM) to provide reliable thermal comfort patterns of cooling PECS from limited experimental datasets. The decision‐making framework integrates a personal characteristic‐based rule refinement and prioritization logic to reflect individual traits to effective cooling modalities. In particular, the proposed framework can simultaneously optimize several thermal comfort metrics and effectively capture the interplay between occupant‐specific attributes and localized thermal interventions of cooling PECS devices to provide actionable PECS selection guideline. The results of this study offer a data‐efficient pathway for addressing heterogeneous cooling demands and advancing human‐centric building climate control using cooling PECS.