基于人工智能的建筑能耗预测研究综述  

Review of Building Energy Consumption Prediction Based on Artificial Intelligence

作　　者：冯增喜[1] 杨芸芸 赵锦彤 何鑫 张茂强 崔巍[1] 王泽 FENG Zengxi;YANG Yunyun;ZHAO Jintong;HE Xin;ZHANG Maoqiang;CUI Wei;WANG Ze(School of Building Services Science and Engineering,Xi’an University of Architecture and Technology,Xi’an 710055,China;School of Metallurgical Engineering,Xi’an University of Architecture and Technology,Xi’an 710055,China)

机构地区：[1]西安建筑科技大学建筑设备科学与工程学院,西安710055 [2]西安建筑科技大学冶金工程学院,西安710055

出　　处：《建筑节能（中英文）》2023年第3期22-29,共8页Building Energy Efficiency

基　　金：安徽建筑大学智能建筑与建筑节能安徽省重点实验室2018年度开放课题(IBES2018KF08)。

摘　　要：能耗预测作为建筑节能的先决条件,对于挖掘建筑节能潜力、提高建筑设备使用效率、提升建筑运营管理质量有重要意义。针对能耗预测,从输入数据类型、输出数据类型、预测时间范围、预测方法4个方面对基于人工智能方法的建筑能耗预测应用现状进行研究。又介绍了单一预测模型和集成预测模型的基本理论,并分别以多元线性回归方法、人工神经网络和支持向量机3种人工智能方法作为基本模型,进行了基于单一预测模型和集成预测模型的建筑能耗预测研究。研究表明与单一预测模型相比,集成预测模型具有更好的预测精度、稳定性及多样性,同时,对人工智能方法在建筑能耗预测领域的应用前景进行了展望。As a prerequisite of building energy conservation,energy consumption prediction is of great significance to excavate building energy conservation potential,improve the efficiency of construction equipment and improve the quality of building operation and management.Aiming at energy consumption prediction,firstly,the application status of building energy consumption prediction based on artificial intelligence method is studied from four aspects:input data type,output data type,prediction time range and prediction method.Then,the basic theories of single prediction model and ensemble prediction model are introduced,and three artificial intelligence methods,including multiple linear regression method,artificial neural network and support vector machine,are used as the basic models to predict building energy consumption based on single prediction model and ensemble prediction model.The research shows that compared with the single prediction model,the ensemble prediction model has better prediction accuracy,stability and diversity.Meanwhile,the application prospect of artificial intelligence method in the field of building energy consumption prediction is forecasted.

关 键 词：人工智能方法 集成预测模型 多元线性回归方法 人工神经网络 支持向量机 

分 类 号：TU111[建筑科学—建筑理论]