Transient pattern analysis for fault detection and diagnosis of HVAC systems

被引：41

作者：

Cho, SH

Yang, HC

Zaheer-uddin, M ^{[1
]}

Ahn, BC

机构：

[1] Concordia Univ, Dept Bldg Civil & Environm Engn, Montreal, PQ H3G 1M8, Canada

[2] Jeonju Univ, Dept Mech & Automot Engn, Jellabukdo 520759, South Korea

[3] KIER, Bldg Energy Res Ctr, Taejon 305343, South Korea

[4] Kyung Won Univ, Dept Bldg Equipment Syst Engn, Kyung Kee Do 461701, South Korea

来源：

ENERGY CONVERSION AND MANAGEMENT | 2005年 / 46卷 / 18-19期

关键词：

FDD (fault detection and diagnosis); VAV (variable air volume); controller gain; transient pattern analysis; HVAC systems;

D O I：

10.1016/j.enconman.2005.02.012

中图分类号：

O414.1 [热力学];

学科分类号：

摘要：

Modern building HVAC systems are complex and consist of a large number of interconnected subsystems and components. In the event of a fault, it becomes very difficult for the operator to locate and isolate the faulty component in such large systems using conventional fault detection methods. In this study, transient pattern analysis is explored as a tool for fault detection and diagnosis of an HVAC system. Several tests involving different fault replications were conducted in an environmental chamber test facility. The results show that the evolution of fault residuals forms clear and distinct patterns that can be used to isolate faults. It was found that the time needed to reach steady state for a typical building HVAC system is at least 50-60 min. This means incorrect diagnosis of faults can happen during online monitoring if the transient pattern responses are not considered in the fault detection and diagnosis analysis. (c) 2005 Elsevier Ltd. All rights reserved.

引用

页码：3103 / 3116

页数：14

共 7 条

[1]

[Anonymous], 1999, ASHRAE Transactions

[2] DETECTING CHANGES IN SIGNALS AND SYSTEMS - A SURVEY [J].