基于回归模型的太阳能热水系统两种典型故障检测与诊断  

Two Kinds of Typical Fault Detection and Diagnosis of Solar Water-Heating System Based on the Regression Model

作　　者：阴元博 高岩[1,2] 张磊[3] 郭海豹[1,2] 王斌[1,2] 

机构地区：[1]北京建筑大学,北京市供热供燃气通风及空调工程重点实验室,北京100044 [2]北京建筑大学,北京市未来城市设计高精尖创新中心,北京100044 [3]中国建筑科学研究院,北京100013

出　　处：《湘潭大学自然科学学报》2017年第3期56-60,共5页Natural Science Journal of Xiangtan University

基　　金：国家自然科学基金项目(51678024)

摘　　要：针对太阳能热水系统常见故障,提出了基于数据回归模型太阳能热水系统故障检测诊断的方法.以对系统进行故障分析为基础,根据残差变化趋势以及与阈值的关系对太阳能热水系统进行故障检测与诊断.通过测得的故障数据,对模型进行验证.数据分析显示:在积灰故障下,集热器出口温度残差达7℃,逐时效率残差超过了10%,逐时得热量参加超过了10kJ;在漏水状态下,集热器出口温度残差达10℃,逐时效率残差超过了50%,逐时得热量参加超过了100kJ,而且温度残差有很大幅度的波动.与理论分析结合可以准确地检测出故障,证明了该方法的有效性.As for common faul ts for solar water heating s ystem, the paper refers a way of faul t detec t ion and diagnosis of solar water-heat ing system based on the regression model. Based on the faul t analysis of system, according to the residual change t rend and the relat ionship wi th the threshold, faul t detect ion and diagnosis was done for solar hot water system. Through the measured faul t data, to veri fy this model. Data analysis showed that： underdepos ition fault , the residual of collector out let temp e ra tu re is 7 °C , th e hour ly ef f iciency residual more than 10 percent , the hourly heat gain in more than 10 kj; In the condi t ion of leakage faul t, the residual of col lector out let temperature is 10 °C , the hourly ef f iciency residual more than 50 percent , the hourly heat gain in more than 100 kj, and residual temperature has a great range of f luctuat ion. Combined wi th theoret ical analysis can accurately detect the faul t, prove the effectiveness of the proposed method.

关 键 词：回归模型 太阳能热水系统 故障检测 故障诊断 故障分析 

分 类 号：TK515[动力工程及工程热物理—热能工程]