摘要
老年人有明显不同于其他年龄段人员的特点,不宜直接采用传统热舒适PMV预测模型进行评价。基于节点分裂算法优化以及决策树赋权的改进随机森林,提出一种老年人热舒适仿真理论模型。以杭州地区为例,通过现场测试、问卷调查及访谈方式构建老年人热舒适仿真数据库。采用笔者模型对杭州市老年人热舒适进行仿真,仿真准确率达82.25%。基于模型训练过程为老年人热舒适影响因素赋予量化权重,研究结果能为老年人热舒适评价和老年人建筑热环境设计提供参考。
The traditional thermal comfort assessment of the PMV method is not suitable for elderly people since they are different from people at other age groups.A simulation model is proposed to investigate the thermal comfort of elderly people.The method of random forest is optimized by combining the note split algorithm with the weighted decision tree algorithm.A thermal comfort database of elderly people is constructed by field measurements and questionnaires as well as interviews in Hangzhou city.The proposed model is trained by the constructed database.The accuracy of the proposed model is up to 82.25%.The indicators of thermal comfort assessment of elderly people are assigned quantitative weights by the training process of the proposed model.This study provides a reference for the thermal comfort assessment of elderly people and the thermal environment design of buildings.
作者
杨玉兰
李洋
邰惠鑫
刘抚英
倪彤元
YANG Yulan;LI Yang;TAI Huixin;LIU Fuying;NI Tongyuan(School of Design and Architecture,Zhejiang University of Technology,Hangzhou 310023,China)
出处
《浙江工业大学学报》
CAS
北大核心
2021年第1期47-52,共6页
Journal of Zhejiang University of Technology
基金
浙江省自然科学基金资助项目(LY18E080035)
国家自然科学基金资助项目(51878125)
杭州市社会发展科研专项计划项目(20130533B19)。
关键词
老年人
热舒适
随机森林
权重
elderly people
thermal comfort
random forest
weight
