基于k-means聚类的动态换热站代表室温提取方法研究  

Research on the Method of Extracting Representative Room Temperatures for Dynamic Heating Stations Based on K-means Clustering

作　　者：相明宇 周恩泽[1] 杨英帅 于君利 徐文晓 刘亚宁 杜岷佳 XIANG Mingyu;ZHOU Enze;YANG Yingshuai;YU Junli;XU Wenxiao;LIU Yaning;DU Minjia(Qingdao University of Technology,Qingdao 266520,Shandong,China;Diehl Metering(Jinan)Co.,Ltd.,Jinan 250101,Shandong,China)

机构地区：[1]青岛理工大学,山东青岛266520 [2]代傲表计(济南)有限公司,山东济南250101

出　　处：《建筑科学》2025年第2期237-245,共9页Building Science

基　　金：中国政府/世界银行/全球环境基金——中国可再生能源规模化发展二期资助项目(QUT-2017-ZX-0010)。

摘　　要：室温传感器的普及使得供热公司可以获得大量的供暖用户室温数据。然而,面对千差万别的用户室温数据,一直缺乏有效的手段来提取能够全面反映换热站供暖效果并为换热站运行调控提供反馈的换热站代表室温。本文基于k-means聚类分析提出了1种动态换热站代表室温提取方法,通过多个代表用户来全面反映绝大部分用户的室温水平与波动特征,并不断动态更新代表用户,实现了整个采暖季上的更准确的换热站供暖效果评估和反馈控制。将所提方法应用于洛阳市某小区换热站,成功提取了换热站代表室温,验证了所提方法的可行性与稳定性,并统计分析了不同换热站调节工况下代表用户代表性持续时长。同时,讨论分析了换热站代表室温的新的应用思路。结果证明,所提方法满足实际工程的应用需要,对于促进供热系统的智能化发展具有积极作用。The popularization of room temperature sensors allows heating companies to obtain a large amount of room temperature data from heating users.However,in the face of the widely varying user room temperature data,there has been a lack of effective means to extract the representative room temperature of the heat exchanger station that can comprehensively reflect the heating effect of the heat exchanger station and provide feedback for the operation and control of the heat exchanger station.In this paper,a dynamic heat transfer station representative room temperature extraction method was proposed based on k-means cluster analysis,which comprehensively reflects the room temperature level and fluctuation characteristics of the majority of users through multiple representative users,and dynamically updates the representative users,realizing a more accurate assessment of the heating effect of the heat transfer station and feedback control in the whole heating season.The proposed method was applied to a district heat exchanger station in Luoyang,and the representative room temperature of the heat exchanger station was successfully extracted,to verify its feasibility and stability.Moreover,a statistical analysis was conducted on the representative duration of the representative users under different regulation conditions of the heat exchanger station.Meanwhile,a new idea of applying the representative room temperature of a heat exchanger station was discussed and analyzed.The results proved that the proposed method can meet the application needs of practical projects and plays a positive role in promoting the intelligent development of heating systems.

关 键 词：换热站代表室温 K-MEANS 室温反馈 智慧供热 

分 类 号：TU111.191[建筑科学—建筑理论]