中文摘要：

人体在局部微环境与背景环境之间过渡时,热感觉是随着环境参数的变化而发生相应变化的。以往的各种方法总是试图寻找热感觉的精确预测,但由于热反应本身的复杂性,建立的模型往往不是很理想。本文提出一种基于支持向量机（Support Vector Machine,SVM）的热感觉自动分类方法,将热感觉的预测转化成模式分类问题。通过提取环境空间中的6个参数及其不同组合方式,对不同过渡方案后的热感觉进行分类,经验证此分类模型能够取得较好的分类效果。结果还表明工位区与背景区的黑球温度、工位区的送风速度对分类效果基本不产生影响,背景区温度、工位区的送风温度、过渡后的时间对分类效果影响较大。

英文摘要：

When the body transit between the local micro environment and the background environment,thermal sensation will change with the change of environment parameter.The previous methods always attempt to seek for the exact prediction of thermal sensation,but because of the complexity of thermal response,it is not very ideal for the establishment model.The article propose a thermal sensation automatic classification method based on SVM,which is not aimed at pursuing exact prediction,but converted into a question of pattern classification.By extracting the 6 parameters of environmental space and their different combinations,it is sorted for the thermal sensation of the different transition program.It is validated that the classification model can achieve better classification results.The results also show that it does not affect the classification results for the black globe temperature of the station area and background area、air speed of station area.The impact on classification results is greater for the background temperature、supply air temperature of station area and the elapsed time after transition.