摘要
虽然大型综合医院已经普遍通过加装智能电表实现用电回路监测,但是医院建筑节能仍多采用节能改造等方法,传统节能针对数据利用能力普遍较弱。针对上述问题,提出了一种基于建筑信息模型(BIM)的医院建筑节能诊断方法。首先,建立了医院能耗监测网络,在建筑信息模型中整合用电数据、用电外部影响因子;然后,基于k-means聚类算法和离群点检测两项无监督学习算法实现异常用电数据的主动挖掘;最后,基于建筑信息模型可视化展示建筑内用电逻辑、用电分布、异常点位置。在实际工程项目应用中验证了所提方法的有效性,可从可视化、智能化等角度从丰富的医院建筑用电数据中发现异常现象和节能潜力点,辅助节能诊断和管理。
Although large general hospitals have generally implemented electricity circuit monitoring by installing smart meters,hospital building energy conservation still uses energy-saving transformation methods,however traditional energy-saving targets are generally weak in data utilization.Aiming at the above problems,a method for diagnosing energy efficiency of hospital buildings based on Building Information Model(BIM)was proposed.Firstly,a hospital energy consumption monitoring network was established to integrate electricity consumption data and external influence factors of electricity consumption in the building information model.Then,the active mining of abnormal electricity consumption data was achieved based on two unsupervised learning algorithms of k-means clustering algorithm and outlier detection.Finally,based on the building information model,the electricity usage logic,electricity usage distribution,and abnormal point locations in the building were visually displayed.The effectiveness of the proposed method is verified in the application of a real engineering project.The proposed method can find the abnormal phenomena and energy-saving potential points from the rich hospital building electricity data from the perspective of visualization and intelligence,and assist energy-saving diagnosis and management.
作者
许璟琳
彭阳
余芳强
XU Jinglin;PENG Yang;YU Fangqiang(Shanghai Construction No.4(Group)Company Limited,Shanghai 201103,China)
出处
《计算机应用》
CSCD
北大核心
2021年第S01期288-292,共5页
journal of Computer Applications
基金
上海市扬帆计划项目(18YF1410400,19YF1421100)
上海市启明星课题(18QB1402300)。
关键词
医院建筑
节能诊断
K-MEANS聚类
离群点检测
建筑信息模型
hospital building
energy-saving diagnosis
k-means clustering
outlier detection
Building Information Model(BIM)
