基于空调设备能耗模型的半在线辨识方法研究  

Research on half-online parameter estimation based on air-conditioning energy consumption model

作　　者：周孝锋 张迪 石定贤 李斌[1] ZHOU Xiao-feng;ZHANG Di;SHI Ding-xian

机构地区：[1]国防大学联合勤务学院,北京100858 [2]军委后勤保障部工程质量监督中心,北京100036

出　　处：《节能》2018年第9期37-41,共5页Energy Conservation

摘　　要：针对空调设备在使用过程中能耗模型变化的特点,提出了一种采用"在线监测+条件触发辨识"的半在线参数估计算法。通过构建风机性能老化测试数据,与传统的3种参数估计方法进行了应用性能的对比。对半在线参数估计算法中3个参数(截取的数据窗口长度L、相对误差阈值h和连续上限次数阈值M)的选取准则进行了研究。Concerning the salient features of components energy consumption model in air-conditioning system, a new parameter estimation method by means of ＂on-line monitoring ＋ conditional trigger＂, named half-online least square （HOLS） algorithms, was proposed. The comparison of HOLS and three general algorithms （one-time least square estimation LS, general recursive least square estimation RLS and least square estimation with forgetting factor FFRLS） was committed to study the application performance of HOLS. The selection criteria for the three parameters of HOLS, data window size L, relative error threshold h and continuous times M, were studied.

关 键 词：空调系统 半在线参数辨识 能耗模型 

分 类 号：TP271[自动化与计算机技术—检测技术与自动化装置] TU831.6[自动化与计算机技术—控制科学与工程]