商场空间形式的热性能预测模型

Thermal Performance Prediction Model of Shopping Mall Space Form

入口-过渡空间作为入口侵入风在室内传导的必经空间,其空间形态会影响建筑内部人员活动区的温度分布,并从空间温度场均匀度和温度的数值两个角度影响人员热舒适。通过大样本信息收集对寒冷地区大型商场的典型模型和关键影响参量进行提取,并通过CFD仿真计算,探究不同入口-过渡空间形态的大型商场室内热环境温度分布差异。提出了基于空间抗侵入风能力、步行街热舒适和综合空间性能评分的评价指标Tr、Cr和S,分析得到如下结果:现有"入口-过渡空间"形式仍有较大的优化潜力值得挖掘;在大部分情况下,"过渡空间宽度"、"入口与主导风向夹角"和"过渡空间轴线形状"是影响程度最大的3个参量;相较于单个参量的因子权重,空间参量组合产生的耦合作用也非常重要。在此基础上,提出了不同条件下的优势空间形式和劣势空间形式,并采用BP神经网络算法训练模拟样本,建立了空间性能预测模型,为建筑师空间设计提供建筑性能辅助决策工具。

The entrance-transition space as the necessary space for the entrance intrusive wind conduction,its spatial form will affect the temperature distribution of occupant zone inside the building,and affect the thermal comfort of the personnel from two aspects of space temperature uniformity and temperature value.The typical model and key parameters of shopping malls in cold regions were extracted by large sample information collection,and the difference of indoor temperature distribution with different entrance-transition space forms was explored through CFD simulation calculation.We proposed the evaluation indexes Tr,Cr and S,based on the scores of space anti-invasion wind ability,thermal comfort of pedestrian street and comprehensive space performance respectively.The results showed that:the existing"entrance-transition space"form still has a large optimization potential worth exploring;in most cases,"transition space width","entrance wind direction"and"transition space axis shape"are the three most influential parameters;compared with the factor weight of a single parameter,the coupling effect of spatial parameter combination is also very important.On this basis,we proposed the strong forms and weak forms under different conditions,and used the BP neural network algorithm to train the simulation samples,established a space performance prediction model,which provides the space performance auxiliary tool for architect in architectural design.