单位面积供热能耗的家庭建筑面积效应  

Effect of Household Floor Area on Heating Energy Consumption per Unit Area

作　　者：那威[1] 田登港 NA Wei;TIAN Denggang

机构地区：[1]北京建筑大学,北京100044

出　　处：《煤气与热力》2025年第2期18-23,共6页Gas & Heat

基　　金：北京市自然科学基金(9222009)。

摘　　要：基于统计年鉴数据提出了城市住宅供热能耗核算方法,建立了基于STIRPAT模型的单位面积供热能耗代理模型(简称代理模型)。采用统计年鉴住宅供热能耗数据样本,训练获取代理模型的标准化系数。经测试集样本检验,代理模型均方根误差为1.99 kg/m^(2),均方根误差变异系数为16.37%,归一化平均偏差为3.36%,证明代理模型预测精度高、泛化性能强。2012年至2020年,若忽视家庭建筑面积效应对单位面积供热能耗降低产生的促进作用,将导致单位面积供热能耗下降水平被高估4.07%。建议当前家庭建筑面积较小的地区优先采用以家庭建筑面积为单位的供热能耗强度评价指标,该指标可有效避免家庭建筑面积效应对单位面积供热能耗下降产生的促进作用。Based on the statistical yearbook data,the calculation method of urban residential heating energy consumption is put forward,and the surrogate model for heating energy consumption per unit area(referred to as the surrogate model)based on STIRPAT model is established.The sample of residential heating energy consumption data from the statistical yearbook is used to train and obtain the standardized coefficients of the surrogate model.After being tested with a validation set sample,the root mean square error(RMSE)of the surrogate model is 1.99 kg/m^(2),the coefficient of varia-tion of the root mean square error is 16.37%,the nor-malized average deviation is 3.36%,which proves that the surrogate model has high prediction accuracy and strong generalization performance.From 2012 to 2020,if the promotion effect of household floor area on the re-duction of heating energy consumption per unit area is ignored,the reduction level of heating energy consump-tion per unit area will be overestimated by 4.07%.It is recommended that areas with small household building areas prioritize the use of heating energy intensity evaluation indicators based on household building area.This indicator can effectively avoid the promoting effect of household building area on the decrease of heating energy consumption per unit area.

关 键 词：单位面积供热能耗 STIRPAT模型 家庭建筑面积效应 

分 类 号：TU832.02[建筑科学—供热、供燃气、通风及空调工程]