基于回归分析的住宅冷热量需求预测模型研究  

A Prediction Model of Heating and Cooling Demand of Residences Based on Regression Analysis

作　　者：房涛[1] 李洁[2] 王崇杰[1] 尹红梅 FANG Tao;LI Jie;WANG Chongjie;YIN Hongmei(School of Architecture and Urban Planning,Shandong Jianzhu University,Jinan 250101,Shandong,China;School of Architecture,Tianjin University,Tianjin 300012,China)

机构地区：[1]山东建筑大学建筑城规学院,山东济南250101 [2]天津大学建筑学院,天津300072

出　　处：《建筑科学》2019年第12期69-75,共7页Building Science

基　　金：国家自然科学基金青年基金项目“寒冷地区近零能耗住宅设计参数量化控制与关键变量协调机理研究”(51608308);住房城乡建设部2017年科学技术项目计划“寒冷地区近零能耗农宅原型设计与技术保障体系研究”(2017-k1-024);天津市应用基础与前沿技术计划项目“基于动态信息模型的建筑生命周期可持续性评价”(15JCYBJC49000)

摘　　要：现有动态、静态能耗计算方法在建筑方案设计阶段应用的局限性,导致建筑节能计算及优化环节缺失。建立一种适用于方案阶段节能计算及优化的能量需求预测模型,是建筑设计方法创新的重要内容。本文以寒冷地区城镇近零能耗住宅的采暖制冷需求为目标,首先通过单因素敏感性分析筛选出10个关键设计参数,然后通过正交试验设计进行能量需求仿真计算,依据计算结果进行多重线性回归、建立多因素耦合作用下“准动态”能量需求预测模型,并通过实际案例验证了该模型的可靠性。最后,采用Java Script脚本语言开发了DOEP软件,供建筑师在方案设计阶段有效预测住宅采暖制冷需求,以此优化节能设计参数。The limitations of the existing dynamic and steady energy calculations in the design of the construction scheme result in the absence of energy performance calculation and optimization.Establishing a energy demand predicting model that is applicable to calculation and optimization of energy conservation at the scheme stage is critical for innovating the building design method.In this paper,the heating and cooling demands for nearly zeroenergy home in cold areas were emphasized to firstly screen 10 key parameters via sensitivity analysis of single factor.Secondly,energy demand simulation calculation was conducted through the orthogonal test design.Next,after conducting multiple linear regression on the calculation result,the quasi-dynamic energy demand calculation predicting model was built under the coupling effect of multiple factors and verified by a practical case.Finally,the DOEP software was developed with the JavaScript,which can facilitate architects to effectively predict the heating and cooling demeans of residences at the scheme design stage,so that energy conservation design parameters can be optimized.

关 键 词：能耗计算 敏感性分析 正交试验 预测模型 优化参数 

分 类 号：TU831[建筑科学—供热、供燃气、通风及空调工程]