摘要
对寒冷地区民用建筑温度舒适度评价方法的设计,能够更好的改善寒冷地区民用建筑冬季室内的环境状况。对室内温度舒适度评价方法进行设计时,由于采集因素不够全面,导致得到的综合评价指标不够完整。传统方法先对影响因素做整体分析,得到多元舒适性综合评价指标体系,但没有给出最终评价规则,导致评价结果不准确。提出基于支持向量机的寒冷地区民用建筑温度舒适度评价方法。采集影响寒冷地区民用建筑温度舒适度多种因素,组建多元舒适性综合评价指标体系,融合于支持向量机理论训练分类器,利用分类器得到训练数据集每一模式的所有评价指标累计贡献率分量的最大值,并给出寒冷地区民用建筑温度舒适度评价规则,完成对寒冷地区民用建筑温度舒适度评价。仿真结果表明,所提方法评价精确度高,可以为改善寒冷地区民用建筑冬季室内的环境状况提供有力的依据。
This article proposes a method for temperature comfort evaluation of civil architecture in cold region based on support vector machine (SVM). Firstly, the article collected various factors influencing temperature comfort of civil architecture and built multi component system of comprehensive evaluation index of comfort, and then trained classifier integrated with the SVM and used the classifier to obtain maximum of component of accumulative contribu- tion rate of all evaluation indexes in each mode of train data set. Further, the rule of the temperature comfort evalua- tion was provided. Simulation results show that the method has higher evaluation precision and can provide powerful basis for improving indoor environmental condition of civil architecture in winter.
出处
《计算机仿真》
北大核心
2017年第8期445-448,共4页
Computer Simulation
关键词
寒冷地区
民用建筑
温度舒适度评价
Cold region
Civil architecture
Temperature comfort evaluation