A simulation model for visitors' thermal comfort at urban public squares using non-probabilistic binary-linear classifier through soft-computing methodologies

作     者：Kariminia, Shahab Shamshirband, Shahaboddin Hashim, Roslan Saberi, Ahmadreza Petkovic, Dalibor Roy, Chandrabhushan Motamedi, Shervin 

作者机构：Islamic Azad Univ Najafabad Branch Fac Art Architecture & Urban Planning Dept Architecture Najafabad Isfahan Iran Univ Malaya Fac Comp Sci & Informat Technol Dept Comp Syst & Technol Kuala Lumpur 50603 Malaysia Univ Malaya Fac Engn Dept Civil Engn Kuala Lumpur 50603 Malaysia Univ Malaya Inst Ocean & Earth Sci Kuala Lumpur 50603 Malaysia Univ Teknol MARA Fac Architecture Planning & Surveying Shah Alam 40450 Malaysia Univ Nis Fac Mech Engn Aleksandra Medvedeva 14 Nish 18000 Serbia 

出 版 物：《ENERGY》 (能源杂志)

年 卷 期：2016年第101卷

页      面：568-580页

核心收录：

学科分类：0820[工学-石油与天然气工程] 08[工学] 0807[工学-动力工程及工程热物理] 

基　　金：HIR-MOHE University of Malaya [UM.C/HIR/MOHE/ENG/34] 

主　　题：Thermal comfort conditions Outdoor spaces Support vector machine Wavelet algorithm Microclimate 

摘      要：Sustaining outdoor life in cities is decreasing because of the recent rapid urbanisation without considering climate-responsive urban design concepts. Such inadvertent climatic modifications at the indoor level have imposed considerable demand on the urban energy resources. It is important to provide comfortable ambient climate at open urban squares. Researchers need to predict the comfortable conditions at such outdoor squares. The main objective of this study is predict the visitors outdoor comfort indices by using a developed computational model termed as SVM-WAVELET (Support Vector Machines combined with Discrete Wavelet Transform algorithm). For data collection, the field study was conducted in downtown Isfahan, Iran (51 degrees 41 E, 32 degrees 37 N) with hot and arid summers. Based on different environmental elements, four separate locations were monitored across two public squares. Meteorological data were measured simultaneously by surveying the visitors thermal sensations. According to the subjects thermal feeling and their characteristics, their level of comfort was estimated. Further, the adapted computational model was used to estimate the visitors thermal sensations in terms of thermal comfort indices. The SVM-WAVELET results indicate that R-2 value for input parameters, including Thermal Sensation, PMW (The predicted mean vote), PET (physiologically equivalent temperature), SET (standard effective temperature) and T-mrt were estimated at 0.482, 0.943, 0.988, 0.969 and 0.840, respectively. (C) 2016 Elsevier Ltd. All rights reserved.