基于分项计量数据对两种冷水机组序列启停策略的识别算法

An Algorithm Using Sub-Metering Data to Identify Two Chiller Sequencing Control Strategies

冷水机组的序列控制方式对建筑能耗和室内热舒适有着显著的影响,如能对其进行自动识别和优化,对公共建筑节能具有重要的意义。从2008年开始,我国在全国范围内推行大型公共建筑的分项计量系统,冷热源系统的电耗是该系统的必要监测内容,该系统为自动识别冷水机组的启停策略提供了数据基础。因而,基于冷水机组的电耗数据,利用数据挖掘的算法来识别冷水机组序列启停方式的方法。该方法利用One R分类算法来实现对以建筑即时冷负荷为指标的冷水机组序列控制方式(Q-based)和以压缩机功率为指标的控制方式(P-based)这两者的区分。最后,通过对位于上海市的一栋建筑进行模拟实验来验证该算法的可行性。

For a multiple-chiller plant, chiller sequencing control significantly affects both building energy consumption and indoor temperature controls （indoor thermal comfort）, hence the identification of chiller sequencing control is of great importance to energy saving of public buildings. Since 2008, the submetering systems of large public building have been applied for real-time monitoring and controlling in China and national level. To fill the gap, this paper proposes a method to identify the exact chiller sequencing control strategy and its related parameters. The proposed method uses a data-mining algorithms- OneR, to analyze the sub-metering data, aiming to discover the chiller sequencing control strategies. Finally, simulation experiments of several buildings located in Shanghai were conducted to validate the feasibility of the proposed method.