1、 International Journal of Heating, Ventilating, Air-conditioning and Refrigerating Research Editor Reinhard Radermacher, Ph.D., Professor and Director, Center for Environmental Energy Engineering, Department of Mechanical Engineering, University of Maryland, College Park, USA Associate Editors James
2、 E. Braun, Ph.D., P.E., Professor, Ray W. Herrick Laboratories, Alberto Cavallini, Ph.D., Professor, Dipartmento di Fisicia Tecnica, UniversiQ of Padova, Italy Qingyan (Yan) Chen, Ph.D., Professor of Mechanical Engineering, School of Mechanical Engineering, Purdue University, West Lafayette, Indiana
3、, USA Arthur L. Dexter, D.Phil., C.Eng., Professor of Engineering Science, Department of Engineering Science, University of Oxford, United Kingdom Srinivas Garimella, Ph.D., Associate Professor and Director, Advanced Thermal Systems Laboratory, Department of Mechanical Engineering, Iowa State Univer
4、sity, Ames, Iowa, USA Leon R. Glicksman, Ph.D., Professor, Departments of Architecture and Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, USA Anthony M. Jacobi, Ph.D., Professor and Co-Director ACRC, Department of Mechanical and Industrial Engineering, University of Illino
5、is, Urbana-Champaign, USA Bjarne W. Olesen, Ph.D., Professor, Intemational Centre for Indoor Environment and Energy Technical University of Denmark Nils Koppels All, Lyngby, Denmark Jeffrey D. Spitler, Ph.D., P.E., Professor, School of Mechanical and Aerospace Engineering, Oklahoma State University,
6、 Stillwater, Oklahoma, USA School of Mechanical Engineering, Purdue University, West Lafayette, Indiana, USA Editorial Assistant Lon Puente, CEEE OfficeMechanical Engineering, University of Maryland (301 -405-5439) Policy Committee Special Publications Staff Daryl Boyce, Chair, Member ASHRAE P. Ole
7、Fanger, Fellow/Li$e Member ASHRAE Curtis O. Pedersen, Fellow ASHRAE Reinhard Radermacher, Member ASHRAE Jeff Littleton, Associate Member ASHRAE W. Stephen Comstock, AssociateMember ASHRAE Mildred Geshwiler, Editor Erin S. Howard, Associate Editor Christina Helms, Associate Editor Michshell Phillips,
8、 Secretary W. Stephen Comstock Publisher 02005 by the American Society of Heating,.Refrigerating and Air- Conditioning Engineers, Inc., 1791 Tullie Circle, Atlanta, Georgia 30329. All rights reserved. Periodicals postage paid at Atlanta, Georgia, and additional mailing offices. HVAC nor may any pari
9、 of this book be reproduced, stored in a retrieval system, or transmitted in any fnrm or by any rneans-electronic. photocopying, recording, or other-without permission in writing from ASHRAE. Abstracts-Abstracted and indexed by ASHRAE Abstract Center; Ei (Engineering Information, Inc.) Ei Compendex
10、and Engineering Index; IS1 (Institute for Scientific Information) Weh Science and Research Alert; and BSNA (Building Services Research accepted July 20,2004 Part I of this article was published in Volume I I, Number I, January 2005. This paper is the second of a two-part review of methodsfor automat
11、ed fault detection and diag- nostics (FDD) andprognostics whose intent is to increase awareness of the HVACM research and development community to the body of FDD and prognostics developments in other fields as well as advancements in theJield of HVAC Fasolo and Seborg (1995) for HVAC Li et al. (199
12、6, 1997) for heating systems; Isermann and Nold (1988) and Dalton et al. (1995) for pumps; Noura et al. (1993) for large thermal plants; Isermann and Ball (1997) for applications for motors; and Dodier and Kreider (1999) for whole-building systems. Refrigerators One of the early applications of FDD
13、was to vapor-compression-cycle-based refrigerators (McKellar 1987; Stallard 1989). Although McKellar (1987) did not develop an FDD system, he identified common faults for a refrigerator based on the vapor-compression cycle and investi- gated the effects of the faults on the thermodynamic states at v
14、arious points in the cycle. He con- cluded that the suction pressure (or temperature), discharge pressure (or temperature), and the discharge-to-suction pressure ratio were sufficient for developing an FDD system. The faults considered were compressor valve leakage, fan faults (condenser and evapora
15、tor), evaporator frosting, partially blocked capillary tubes, and improper refrigerant charge (under and over charge). Building upon McKellars work, Stallard (1989) developed an automated FDD system for refrigerators. A rule-based expert system was used with simple limit checks for both detection an
16、d diagnosis. Condensing temperature, evaporating temperature, condenser inlet temperature, and the ratio of discharge-to-suction pressure were used directly as classification features. Faults were detected and diagnosed by comparing the change in the direction of the measured quanti- ties with expec
17、ted values and matching the changes to expected directional changes associated with each fault. Air Conditioners and Heat Pumps There are many applications of FDD to air conditioners and heat pumps based on the vapor- compression cycle. Some of these studies are discussed below (Yoshimura and Ito 19
18、89; Kumamaru et al. 1991; Inatsu et al. 1992; Wagner and Shoureshi 1992; Rossi 1995; Rossi and Braun 1996, 1997; Breuker 1997; Breuker and Braun 1998b; Ghiaus 1999; Chen and Braun 2000). Breuker and Braun (1998a) summarized common faults in air .conditioners and their impact on performance. In addit
19、ion, the frequency of fault occurrence and the relative cost of service for various faults were estimated from service records. Yoshimura and lto (1 989) used pressure and temperature measurements to detect problems with condenser, evaporator, compressor, expansion valve, and refrigerant charge on a
20、 packaged air conditioner. The differences between measured values and expected values were used to detect faults. Expected values were estimated from manufacturers data, and the thresholds for fault detection were experimentally determined in the laboratory. Both detection and diagnosis were conduc
21、ted in a single step. No details were provided as to how the thresholds for detection were selected. Wagner and Shoureshi (1 992) developed two different fault detection methods and compared their abilities to detect five different faults in a small heat pump system in the laboratory. The five fault
22、s included abrupt condenser and evaporator fan failures, capillary tube blockage, com- pressor piston leakage, and seal system leakage. The first method was based on limit and trend VOLUME 1 1, NUMBER 2, APRIL 2005 171 Table 1. Symptom Patterns for Selected Faults (Grimmelius et al. 1995) W Compress
23、or, Suction Side,IncreaseinFlow some of the studies are summarized below (Grimmelius et al. 1995; Gordon and Ng 1994, 1995; Stylianou and Nikanpour 1996; Tsutsui and Kamimura 1996; Peits- man and Bakker 1996; Stylianou 1997; Bailey 1998; Sreedharan and Haves 2001; Castro 2002). Comstock et al. (1999
24、) and Reddy et al. (2001) provide a detailed review of FDD literature relat- ing to chiller systems up to their respective times. Cornstock et al. (2002) presented a list of common chiller faults and their impacts on performance. Grimmelius et al. (1995) developed a fault diagnostic system for a chi
25、ller, in which fault detection and diagnostics are carried out in a single step. The FDD method uses a reference model based on multivariate linear regression that was developed with data from a properly operating chiller to estimate values for process variables for a healthy (unfaulted) chiller. Th
26、ese estimates are subsequently used to generate residuals (Le., differences between actual measured values and the values from the reference model). Patterns of these residuals are compared to characteristic patterns corresponding to faulted conditions, and scores are assigned indicating the degree
27、to which the patterns match the pattern corresponding to each fault mode. Fault modes with good fits (high scores) are judged as probably existing in the chiller. Fault modes with poor fits (low scores) are judged as unlikely to exist in the chiller, and faults with interme- diate scores are labeled
28、 as possibly existing. Twenty different measurements are used including VOLUME 11, NUMBER 2, APRIL 2005 173 Table 2. Scoring of Fault Modes for a Highly Idealized Example Fault Mode/ Symptom Symptom Symptom Symptom Total Score Normalized Score 1 2 3 4 Score FI .1 t Scores 10 10 10 10 40 1 .o F2 T -+
29、 T + Scores O 9 O 3 12 0.3 Measurement- 4 -+ .1 t Based Pattern temperatures, pressures, power consumption, and compressor oil level. In addition to the mea- sured variables, some derived variables, such as liquid subcooling, superheat, and pressure drop, are used. The inputs to the model also inclu
30、de the outdoor ambient temperature and load condi- tions. To identie potential fault modes, the chiller is classified into seven components: compressor, condenser, evaporator, expansion valve, liquid line immediately downstream of the condenser and including a filter drier, liquid line with solenoid
31、 and sight glass between the other liquid line and the evaporator, and the crankcase heater. Fault modes are associated with any component that is serviceable, which leads to 58 different fault modes. A cause and effect study of the 58 fault modes helped establish the expected influence of the fault
32、s on the components, measured variables, and subsequent chiller behavior. Symptoms are defined as a difference in any mea- sured or derived variable from its expected value for normal unfaulted operation (i.e., the value given by the reference model). Symptoms associated with all 58 fault modes were
33、 generated and arranged into symptom patterns. Fault modes having identical symptom patterns were aggre- gated into a single fault mode, reducing the total number of fault modes from 58 to 37. These symptom patterns are arranged in a symptom matrix as shown in Table 2, with each row giving the sympt
34、om pattern associated with a particular fault. A symptom (cell in the matrix) shown by an arrow pointing up, ?, indicates a value for the variable greater than that given by the refer- ence model. Likewise, an arrow pointing down, however, only one example (air in the system) is described in their p
35、aper. ANN models appeared to have a slightly better perfor- mance than the ARX models in detecting faults at both the system and the component levels. The authors also note that it is critical to find a global minimum when using ANN models. If an incorrect initial state is chosen, it may lead to a l
36、ocal minimum rather than the global minimum. Bailey (1998) also used an ANN model to detect and diagnose faults in an air-cooled chiller with a screw compressor. The detection and diagnosis were carried out in a single step. The faults evaluated included refrigerant under- and overcharge, oil under-
37、 and overcharge, con- denser fan loss (total failure), and condenser fouling. The measured data included superheat for heat exchanger circuits 1 and 2, subcooling from circuits 1 and 2, power consumption, suction pressure for circuits 1 and 2, discharge pressures for circuits 1 and 2, chilled water
38、inlet and out- let temperatures from the evaporator, and chiller capacity. Each heat exchanger circuit has its own compressor. The ANN model was applied to normal and “faulty” test data collected from a 70-ton laboratory air-cooled chiller with screw compressor. Sreedharan and Haves (2001) compared
39、three chiller models for their ability to reproduce the observed performance of a centrifugal chiller. Although the evaluation was meant to find the most suitable model for chiller FDD, no FDD system was proposed or developed. Two models were based on first principles (from Gordon and Ng I9951 and a
40、 modified ASHRAE Primay Toolkit from Bourdouxhe et al. 1997) and the third was an empirical model. While each model has some distinct advantages and disadvantages, they concluded that the accuracies of all three models were similar. Hydeman et al. (2002) reported that the three models compared by Sr
41、eedharan and Haves (2001) were not accurate in predicting the power consumption of chillers with variable condenser water flow and centrifugal chillers operating with variable-speed drives at low loads. They reformulated the Gordon and Ng model and found that it performed better than the three model
42、s described above. Castro (2002) used a physical model developed by Rossi (1995) along with a k-nearest neigh- bor classifier to detect faults and a rule base to diagnose five different faults (condenser and evaporator fouling, liquid line restriction, and refrigerant under- and overcharge) in a rec
43、iprocat- ing chiller. The FDD implementation detected and diagnosed condenser fouling, refrigerant undercharge at faults level of 20% or greater, and evaporator fouling and liquid line restriction at fault levels of 30% or greater. Refer to Box and Jenkins (i 976) for more details on ARX type models
44、. VOLUME 11, NUMBER 2, APRIL 2005 111 Air-Handling Units There are several studies relating to FDD methods for air-handling units (both the airside and the waterside); some of these are summarized in this section (Norford and Little 1993; Glass et al. 1995; Yoshida et al. 1996; Haves et al. 1996; Le
45、e et al. 1996a, 1996b; Lee et al. 1997; Peits- man and Soethout 1997; Brmbley et al. 1998; Katipamula et al. 1999; House et al. 1999; Ngo and Dexter 1999; Yoshida and Kumar 1999; Seem et al. 1999; Karki and Karjalainen 1999; Morisot and Marchio 1999; House et al. 2001; Dexter and Ngo 2001; Kumar et
46、al. 2001; Sals- bury and Diamond 2001; Carling 2002; Norford et al. 2002; Wang and Chen 2002; Pakanen and Sundquist 2 003). Norford and Little (1993) classi faults in ventilating systems, consisting of fans, ducts, dampers, heat exchangers, and controls. They then review two forms of steady-state pa
47、rametric models for the electric power used by supply fans and propose a third, that of correlating power with a variable-speed drive control signal. The models are compared based on prediction accu- racy, sensor requirements, and their ability to detect faults. Using the three proposed models, four
48、 different types of faults associated with fan systems are detected: (i) failure to maintain supply air temperature, (2) failure to maintain supply air pres- sure setpoint, (3) increased pressure drop, and (4) malfunction of fan motor coupling to fan and fan controls. Although the paper by Norford a
49、nd Little (1993) lacks details on how the faults were evaluated, error analysis and associated model fits were discussed. The results indicate that ail three models were able to identi at least three of the four faults. The diagnosis of the faults is inferred afer the fault is detected. Glass et al. (1 995) use a qualitative model-based approach to detect faults in an air-handling unit. The method uses outdoor, return, and supply air temperatures and control signals for the cooling coil, heating coil, and the damper system. Although Glass et al
