基于相似工况组合权重的空调能耗预测方法  

Combined Weight Prediction Method of Air-Conditioning Energy Consumption based on Similar Conditions

作　　者：马忠娇 张吉礼 MA Zhong-jiao;ZHANG Ji-li(Faculty of Infrastructure Engineering, Dalian University of Technolog)

机构地区：[1]大连理工大学建设工程学部

出　　处：《建筑热能通风空调》2017年第12期13-18,共6页Building Energy & Environment

基　　金：国家自然科学基金项目(No.51578102;No.51378005);大连市科技计划项目(2015E11SF052)

摘　　要：本文提出一种建筑空调系统能耗预测方法,该方法基于天气类型(晴、多云、阴、雨等)与工作日类型(工作日、节假日)相同条件下,利用相似工况变量(室外空气温度、照明插座用电量)的基值点误差,采用组合权重法求取相似误差,在一定相似误差范围内确定相似日,将相似日的空调能耗作为对比日的预测能耗。该方法将主观权重,客观熵权以及相关系数有效融合确定相似日选取的相似误差,通过某大型办公建筑实测数据算例仿真,预测值与实测值相对误差在5%以内。结果表明,在非突变天气情况下,所提方法可选出相似度较高的相似日,预测精度较高。A method for predicting the energy consumption of air-conditioning systems is proposed in this paper. This method is based on the same weather type （sunny, cloudy, overcast, or rainy） and working day type （workdays or holidays）. By calculating the baseline errors of similar working conditions （outdoor temperature and electricity consumptions of lighting and plug）, combined weight method is applied to determine the similar error. Similar days were set within a certain range of similar error, and the air-conditioning energy consumption in these similar days is regarded as that in the predicted days. In order to determine the similar error of similar day, the method effectively combines the subjective weight, objective entropy weight and the correlation coefficient. According to the simulation example of measured data in an office building, the relative errors between the predicted values and the measured are within 5%. That shows the proposed prediction method is with high of similarity under non-catastrophic weather conditions. forecast accuracy and can select similar days with a high degree

关 键 词：空调能耗预测 相似工况 组合权重 熵权法 

分 类 号：TU111.195[建筑科学—建筑理论] TU83