| 摘要 | 第5-7页 |
| ABSTRACT | 第7-10页 |
| NOMENCALTURE | 第18-21页 |
| CHAPTER 1 INTRODUCTION | 第21-39页 |
| 1.1 Motivations | 第21-23页 |
| 1.2 Literature review | 第23-35页 |
| 1.2.1 Model-based methods | 第25-27页 |
| 1.2.2 Data-driven methods | 第27-30页 |
| 1.2.3 Backgrounds of chiller FDD methods | 第30-35页 |
| 1.3 Aim and purposes | 第35-37页 |
| 1.4 Organization of this thesis | 第37-39页 |
| CHAPTER 2 THE OVERVIEW OF BASIC CHILLER FDD METHOD | 第39-50页 |
| 2.1 Descriptions of centrifugal chiller system | 第39-44页 |
| 2.1.1 Basic refrigeration cycle | 第39-41页 |
| 2.1.2 System mechanism | 第41-44页 |
| 2.2 Online FDD strategy for chiller | 第44-48页 |
| 2.2.1 Exponentially Weighted Moving Average residuals for online fault detection | 第44-46页 |
| 2.2.2 Implementation structure of online FDD strategy | 第46-48页 |
| 2.3 The experiment centrifugal chiller data of ASHRAE 1043-RP | 第48-50页 |
| CHAPTER 3 AN ONLINE FDD STRATEGY FOR CENTRIFUGAL CHILLER SYSTEMS BASED ON RADIAL BASIS FUNCTION | 第50-71页 |
| 3.1 Motivations for development of online chiller FDD strategy based on radial basis function | 第50-51页 |
| 3.2 Feature parameters and FDD rule set | 第51-54页 |
| 3.3 Reference models of feature parameters | 第54-55页 |
| 3.4 Tests and comparisons | 第55-66页 |
| 3.4.1 Development of reference models | 第56-58页 |
| 3.4.2 Fault detection and diagnosis results | 第58页 |
| 3.4.3 Normal conditions (Fault-free cases) | 第58-60页 |
| 3.4.4 Fault detection and diagnosis | 第60-66页 |
| 3.5 Validation of FDD strategy using office building chiller system | 第66-69页 |
| 3.5.1 Chiller system | 第66页 |
| 3.5.2 Validation of reference model | 第66-67页 |
| 3.5.3 Fault detection and diagnosis using operating data | 第67-69页 |
| 3.6 Conclusions | 第69-71页 |
| CHAPTER 4 COMPARATIVE INVESTIGATIONS ON REFERENCE MODELS FOR FAULT DETECTION AND DIAGNOSIS IN CENTRIFUGAL Chiller SYSTEMS | 第71-100页 |
| 4.1 Related works | 第71-73页 |
| 4.2 Reference models | 第73-77页 |
| 4.2.1 Multiple linear regression model (MLR) | 第73-74页 |
| 4.2.2 Kriging model | 第74-75页 |
| 4.2.3 Radial basis function model | 第75-77页 |
| 4.3 FDD strategy for chillers | 第77-81页 |
| 4.3.1 Experimental data | 第77-78页 |
| 4.3.2 The sensitivity for fault diagnosis rules | 第78-81页 |
| 4.4 Tests and comparisons | 第81-98页 |
| 4.4.1 Development of the reference models | 第81-84页 |
| 4.4.2 Fault detection and diagnosis results | 第84-96页 |
| 4.4.3 Comparison between two detection methods: EWMA method and t-statistic based method | 第96-98页 |
| 4.5 Conclusions | 第98-100页 |
| CHAPTER 5 AN ENHANCED CHILLER FDD STRATEGY BASED ON DE-LSSVR METHOD COMBINED WITH EWMA | 第100-125页 |
| 5.1 Introduction | 第100-103页 |
| 5.2 The LSSVR chiller model based on the differential evolution method | 第103-111页 |
| 5.2.1 LSSVR basic theory | 第103-105页 |
| 5.2.2 Differential Evolution (DE) | 第105-108页 |
| 5.2.3 LS-SVR parameters optimization based on DE | 第108-111页 |
| 5.3 Validation of the FDD method | 第111-123页 |
| 5.3.1 Development of reference models | 第112-114页 |
| 5.3.2 Fault detection and diagnosis results | 第114-115页 |
| 5.3.3 Normal conditions (Fault-free cases) | 第115-118页 |
| 5.3.4 Fault detection and diagnosis | 第118-123页 |
| 5.4 Conclusions | 第123-125页 |
| CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS | 第125-130页 |
| 6.1 Main contributions | 第125-126页 |
| 6.2 Summary on implementation of the proposed FDD methods | 第126-128页 |
| 6.2.1 An online FDD strategy for centrifugal chiller systems based on radial basis function | 第126-127页 |
| 6.2.2 A FDD strategy for Centrifugal Chiller systems based on comparison studies of chiller models techniques | 第127页 |
| 6.2.3 An enhanced chiller FDD strategy based on LSSVR-DE model combined with EWMA | 第127-128页 |
| 6.3 Recommendations for Future Work | 第128-130页 |
| REFERENCE | 第130-141页 |
| ACKNOWLEDGEMENTS | 第141-142页 |
| PUBLICATIONS | 第142页 |
