Publication: Monitoring Thermal Comfort Level of Commercial Buildings' Occupants in a Hot-Humid Climate Country Using K-nearest Neighbors Model
Date
2020
Authors
Hani Mohamed Salleh F.
Binti Saripuddin M.
Journal Title
Journal ISSN
Volume Title
Publisher
Institute of Electrical and Electronics Engineers Inc.
Abstract
This paper presents a thermal comfort monitoring system that able to simulate the thermal comfort level of commercial building occupants in a hot-humid climate country. The main benefit of this monitoring system is to reduce the energy usage by observing the actual thermal comfort value of the building constantly. The predictive model proposed in this research aims to optimize building energy performance without sacrificing occupant thermal comfort. The system was developed based on KNN (K-nearest neighbors) model using the six thermal comfort factors. The data used for evaluating our proposed model was obtained from a study of adaptive thermal comfort of lecture halls in Malaysia universities. Another set of data is also used to assess the suitability of the proposed model when applied to data of non-hot humid climate (temperate climates). As a result, the proposed monitoring system scores 72.97% of accuracy outperforms PMV that scores 43% accuracy. As for non-hot humid climate country dataset, PMV outperforms our proposed monitoring system, with 35.13% accuracy while the proposed monitoring system only reach up to 18.19%. The results of the experiment reveal that the recommended model is only suitable for use on buildings in hot-humid climate country and may not be suitable for buildings in temperate climate country. � 2020 IEEE.
Description
Climate models; Climatology; Motion compensation; Nearest neighbor search; Office buildings; Predictive analytics; Thermal comfort; Adaptive thermal comfort; Building energy performance; Commercial building; Hot humid climate; K-nearest neighbors; Predictive modeling; Temperate climate; Thermal comfort level; Monitoring