Metabolic rate estimation method using image deep learning
- PDF / 2,587,442 Bytes
- 17 Pages / 612 x 808 pts Page_size
- 25 Downloads / 182 Views
Metabolic rate estimation method using image deep learning Research Article
Hooseung Na, Haneul Choi, Taeyeon Kim () Department of Architectural Engineering, Yonsei University, Seoul 03722, R.O. Korea
Abstract
Keywords
Thermal comfort is an important factor in evaluating indoor environmental quality. However, accurately evaluating thermal comfort conditions is challenging owing to the lack of suitable methods for measuring individual factors such as the metabolic rate (M value). In this study, a M value evaluation method was developed using deep learning. The metabolic equivalent of task was measured for eight typical indoor tasks based on the ASHRAE Standard 55 (lying down, sitting, cooking, walking, eating, house cleaning, folding clothes, and handling 50 kg books) in 31 subjects (males: 16; and females: 15); the measurements were analyzed in terms of gender and body mass index (BMI). The experimental results were assessed using the reliability of the measured data, the M value difference in terms of gender and BMI, and the measurement accuracy. We developed a M value self-evaluation model using artificial intelligence, which achieved an average coefficient of variation (CV) of 12%. A third-party evaluation model was used to evaluate the M value of one subject based on the learning data acquired from the other 30 subjects; this model yielded a low CV of 54%. For high-activity tasks, males generally had higher M values than females, and the higher the BMI was, the higher was the M value. Contrarily, for low-activity tasks, the lower the BMI was, the higher was the M value. The breakthrough M value evaluation method presented herein is expected to improve thermal comfort control.
thermal comfort,
1
Introduction
E-mail: [email protected]
deep learning, gender, body mass index, wearable device
Article History Received: 15 December 2019 Revised: 30 July 2020 Accepted: 06 August 2020 © Tsinghua University Press and Springer-Verlag GmbH Germany, part of Springer Nature 2020
challenging, and expensive equipment is necessary to track and regularly measure subjects (Ravussin et al. 1982; Nieman et al. 2006). Many studies related to TC regarded personal factors as constants. Moon and Kim (2010) developed an advanced thermal control method using an artificial neural network (ANN) for residential buildings. They fixed the M value to 1 met to estimate the TC and fixed the clothing insulation to 1.0 clo and 0.5 clo in the winter and summer, respectively. Chen et al. (2015) developed model predictive control algorithms for PMV control to minimize energy consumption while maintaining TC; they achieved this by fixing the M value to 1.0 met and the clothing insulation to 0.71 clo. Calvino et al. (2004) applied PMV control using a fuzzy adaptive controller, setting the M value and clothing insulation to 1.0 met and 1.0 clo, respectively. Hence, many previous researchers considered the personal factors of occupants as fixed values for evaluation. However, personal factors, such as M value and clothing insulation, are know
Data Loading...
