Exercise and resting periods:Thermal comfort dynamics in gym environments  

作　　者：Ali Berkay Avci Görkem Aybars Balci Tahsin Basaran 

机构地区：[1]Department of Architecture,Faculty of Architecture,Izmir Institute of Technology,Urla,Izmir,Türkiye [2]Department of Coaching Education,Faculty of Sport Sciences,Ege University,Bornova,Izmir,Türkiye

出　　处：《Building Simulation》2024年第9期1557-1578,共22页建筑模拟（英文）

基　　金：supported by the Scientific and Technological Research Council of Turkiye(TUBiTAK)Foundation under grant number122M883.

摘　　要：Physical exercise spaces emerged as popular facilities due to recognizing the significance of physical well-being.This study investigates the relationship among physiological responses,human body energy transfer modes,and indoor environmental conditions in influencing thermal comfort perception within indoor physical exercise space.Seven male participants engaged in a 30 min constant-work-rate cycling exercise and a 20 min resting period in a climatic chamber.The physiological and environmental responses were recorded during the experiments,and the body’s energy transfer modes were calculated using the collected data.The dataset was prepared using the 2 min averages of the collected data and calculated parameters across the experiment phases,including the features of skin temperature,core temperature,skin relative humidity,heart rate,oxygen consumption,body’s heat transfer rates through convection,radiation,evaporation,and respiration,net metabolic heat production rate(metabolic rate minus external work rate),indoor air temperature,indoor relative humidity,air velocity,and radiant temperature.Gradient boosting regressor(GBR)was selected as the analyzing method to estimate predicted mean vote(PMV)and thermal sensation vote(TSV)indices during exercise and resting periods using features determined in the study.Thus,the four GBR models were defined as PMV-Exercise,PMV-Resting,TSV-Exercise,and TSV-Resting.In order to optimize the models’performances,the hyperparameter tuning process was executed using the GridSearchCV method.A permutation feature importance analysis was performed,emphasizing the significance of net metabolic heat production rate(24.2%),radiant temperature(17.0%),and evaporative heat transfer rate(13.1%).According to the results,PMV-Exercise,PMV-Resting,and TSV-Resting GBR models performed better,while TSV-Exercise faced challenges in predicting exercise thermal sensations.Critically,this study addresses the need to understanding the interrelationship among physiological responses,environmental c

关 键 词：physical exercise gradient boosting THERMOREGULATION FITNESS thermal comfort 

分 类 号：TU111[建筑科学—建筑理论]