一种基于模糊因果聚类的室内热舒适预测方法  

Forecast method based on fuzzy causation clustering for indoor thermal comfort

作　　者：徐远清[1] 陈祥光[1] 王丽[1] 朱铁丹[1] 

机构地区：[1]北京理工大学,北京100081

出　　处：《仪器仪表学报》2006年第z1期850-852,共3页Chinese Journal of Scientific Instrument

摘　　要：提出一种用于室内热舒适预测的模糊因果聚类分析方法。在论域内随机产生若干样本点,对合理样本点运用模糊因果聚类方法求出其最佳聚类,提取各类样本的特征构造模糊预测模型。通过对室内热舒适指标的实例分析,说明该方法可用于不同地区的热舒适预测,能解决PMV等指标预测出现的较大偏差等问题。Based on fuzzy causation clustering,a forecast method is presented for indoor thermal comfort.Sample points are produced randomly in problem space.After eliminating the irrationality sample points,the fuzzy causation clustering method is adopted to gain the best clustering results for the rest reasonable sample points.Furthermore,the feature of each class is extracted to structure a fuzzy forecast model.Experiments demonstrate that the proposed method can forecast indoor thermal comfort in different area and can effectively solve deviation problem for the forecast of the PMV index.

关 键 词：模糊因果聚类 热舒适 预测 

分 类 号：TH7-55[机械工程—仪器科学与技术]