Data repository: Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis

This data repository includes results of the original research article “Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis”. In this article, a method to predict annual heat load profiles with a daily resolution for large consumers from industry and commerce is developed. This method is based on a cluster and regression analysis of natural gas load profiles from 797 German consumers, most with a consumption of more than 1.5 GWh/a. The data repository contains plots of all 797 original normalized load profiles and predicted normalized load profiles in the form of time series. The correlation between daily mean ambient temperature and daily normalized natural gas consumption is visualized in additional figures for each load profile. The files in this data repository are sorted by a two-digit numerical code indicating the economy division according to NACE Rev. 2 [1] (see README) and a one-digit numerical code indicating the detected clusters. The dependency of natural gas consumption on working days on mean daily ambient temperature increases from cluster 0 to cluster 3. The cluster CHP includes consumers that were excluded from the analysis, since they operate a combined heat and power plant (CHP). In a plausibility check, additional consumers were excluded from the analysis. These consumers are assigned to the cluster -1. References [1] Eurostat. NACE Rev.2: Statistical classification of economic activities in the European Community. Luxembourg: Office for Official Publications of the European Communities; 2008.

本数据集仓库收录了原创研究论文《年度工商热负荷曲线：基于k-Means聚类（k-Means clustering）与回归分析的建模》的相关研究成果。该论文提出了一种面向大型工商用户、以日分辨率预测年度热负荷曲线的方法，该方法基于对797名德国用户的天然气负荷曲线开展聚类与回归分析，其中多数用户的年天然气消耗量超过1.5 GWh/a。本数据集仓库包含全部797条原始归一化负荷曲线与预测归一化负荷曲线的时间序列图；针对每条负荷曲线，额外附图展示了日平均环境温度与日归一化天然气消耗量之间的相关性。仓库内的文件按照两位数字编码与一位数字编码进行分类：前者代表依据欧盟经济活动统计分类（NACE Rev. 2）[1]（详见README文件）划分的经济行业分类，后者代表检测得到的聚类簇编号。天然气消耗量受工作日与日平均环境温度的影响程度，从聚类簇0到聚类簇3依次递增。聚类簇CHP包含被排除在分析之外的用户，这类用户均运营有热电联产机组（Combined Heat and Power, CHP）。在合理性校验环节中，另有部分用户被排除出分析，这类用户被划归至聚类簇-1。参考文献[1]：欧盟统计局. NACE Rev.2：欧洲共同体经济活动统计分类. 卢森堡：欧洲共同体官方出版物办公室；2008.