Title Page
Abstract
Contents
Chapter 1. Introduction 17
1.1. Research background and objectives 17
1.2. Research contents and innovations 25
Chapter 2. Overview of research on THIC systems 28
2.1. Structure introduction of THIC system 28
2.1.1. Coupled THIC system structure 28
2.1.2. Decoupled THIC system structure 30
2.2. Hybrid desiccant cooling system in dehumidification applications 33
2.2.1. Hybrid liquid desiccant cooling system 33
2.2.2. Hybrid solid desiccant cooling system 34
2.2.3. Low energy regeneration method 35
2.2.4. System configurations with potential for condensing heat recovery 37
2.3. Decoupling operation strategy for THIC systems 39
2.3.1. General temperature control strategy 40
2.3.2. General humidity control strategy 41
2.4. Application of reinforcement learning in HVAC system optimization 42
2.4.1. Single agent algorithm applications 42
2.4.2. Multi-agent algorithm applications 43
2.5. Summary 44
Chapter 3. Performance analysis of HPDC for the decoupled THIC system application 46
3.1. Overview of scheme and procedure 46
3.1.1. System description and simulation conditions 46
3.1.2. Modelling method and calculation processes 52
3.1.3. System control strategy and performance indexes 59
3.2. Comparative analysis on different HPDC systems 62
3.2.1. Indoor environment under different system configuration 62
3.2.2. System thermodynamic process analysis 68
3.2.3. System performance and energy consumption analysis 76
3.3. Field test and control strategy laboratory experiments 78
3.3.1. Field test objectives and plans 79
3.3.2. Experiment system setup and cases 89
3.4. Experiment results and discussion 92
3.4.1. On-site thermal environment in field tests 92
3.4.2. System electricity consumption in field tests 98
3.4.3. System internal air state changes in laboratory experiments 100
3.4.4. Operation control strategy validation 105
3.5. Summary 108
Chapter 4. Proposal of the temperature and humidity independent control system and decoupling operation control strategy 110
4.1. WLDC system design and validation experiments 111
4.1.1. System concept and operation control 111
4.1.2. Validation experiment setup and conditions 115
4.2. Validation experiment results and discussion 122
4.2.1. Operation state of WLDC system 122
4.2.2. Thermal analysis on air/water side of WLDC system 124
4.2.3. Heat source balance analysis 126
4.3. Introduction of high temperature chilled water THIC system using radiant cooling with desiccant dehumidification system 127
4.3.1. System modelling and simulation conditions 128
4.3.2. System decoupling operation control method 142
4.4. Analysis on the decoupling operation control strategy of the THIC system 145
4.4.1. Indoor thermal environment changes under different system operation schemes 146
4.4.2. Effect of indoor setpoints on the system thermal energy matching 153
4.4.3. Analysis on intelligent control of indoor design temperature and humidity 159
4.5. Summary 173
Chapter 5. Heating source system performance and MADDPG optimization in multi-zone applications 176
5.1. Heat source system operation strategy and key parameters design 176
5.2. Performance analysis of the heat source system with different THIC terminal systems 187
5.3. Dynamic thermal energy change of the system over long periods of operation 195
5.4. Global optimization of system performance by MADDPG algorithm 199
5.4.1. Improvement design of MADDPG algorithm for system global optimization 199
5.4.2. Optimization training of system performance and comparative analysis with control group 202
5.5. Summary 207
Chapter 6. Conclusion 210
Bibliography 214
Abstract in Korean 223
〈Table 3.1〉 Summary of the outdoor condition 51
〈Table 3.2〉 Details of the simulation input parameters 51
〈Table 3.3〉 Coefficients of cooling capacity and EIR modifier curve 57
〈Table 3.4〉 Coefficients of the dehumidification performance curve 57
〈Table 3.5〉 The validation condition and result accuracy 57
〈Table 3.6〉 Field test conditions 85
〈Table 3.7〉 Experiment conditions and case number 91
〈Table 3.8〉 Specifications of sensors and DAQ devices 91
〈Table 3.9〉 Air state for each LP case at steady state 101
〈Table 3.10〉 DCOP results of the HPDC-U system compared with other references 105
〈Table 4.1〉 The relative uncertainty of main parameters 122
〈Table 4.2〉 Input types and values in the design phase 131
〈Table 4.3〉 Operation Conditions in the operation phase 131
〈Table 4.4〉 Various UA required by the cooling coil model 131
〈Table 4.5〉 Building related physical property parameters 141
〈Table 4.6〉 System design parameters 142
〈Table 4.7〉 Basic setup of main elements in the MADDPG algorithm 165
〈Table 5.1〉 The key design parameter and coefficient 180
[Figure 2.1] Diagram of a typical engineering application of the coupled THIC system. 30
[Figure 2.2] System configuration of the decoupled THIC system with AHU 32
[Figure 3.1] System structure diagram of HPDC-U system 47
[Figure 3.2] Air treatment process of the HPDC-U system in a psychrometric chart 48
[Figure 3.3] System structure diagram of HPDC-D system 48
[Figure 3.4] Air treatment process of the HPDC-D system in a psychrometric chart 49
[Figure 3.5] Three-dimensional view of the air-conditioned area 49
[Figure 3.6] Outdoor air conditions of three regions during summer. (a) outdoor temperature; (b) outdoor relative humidity; (c) outdoor... 51
[Figure 3.7] Control flow chart of HPDC systems 60
[Figure 3.8] Indoor air and system supply air state when the HPDC-U system operated in Seoul 63
[Figure 3.9] Indoor air and system supply air state when the HPDC-U system operated in Shanghai 64
[Figure 3.10] Indoor air and system supply air state when the HPDC-U system operated in Singapore 65
[Figure 3.11] Indoor air state and outdoor conditions when the HPDC-D system operated in Seoul 66
[Figure 3.12] Indoor air state and outdoor conditions when the HPDC-D system operated in Shanghai 67
[Figure 3.13] Indoor air state and outdoor conditions when the HPDC-D system operated in Singapore 68
[Figure 3.14] Variation of thermal performance of HPDC-U system in Seoul 69
[Figure 3.15] Variation of thermal performance of HPDC-U system in Shanghai 70
[Figure 3.16] Variation of thermal performance of HPDC-U system in Singapore 71
[Figure 3.17] Comparison of system supply air cooling capacity and indoor sensible load under Seoul climate 72
[Figure 3.18] Comparison of system supply air cooling capacity and indoor sensible load under Shanghai climate 73
[Figure 3.19] Comparison of system supply air cooling capacity and indoor sensible load under Singapore climate 74
[Figure 3.20] Variation of fdeh and freg during the HPDC-D system operation in Seoul[이미지참조] 75
[Figure 3.21] Variation of fdeh and freg during the HPDC-D system operation in Shanghai[이미지참조] 75
[Figure 3.22] Variation of fdeh and freg during the HPDC-D system operation in Singapore[이미지참조] 76
[Figure 3.23] Operation time of different mode during the HPDC-U system operated under three climates 77
[Figure 3.24] Comparison of system MRC in three regions 78
[Figure 3.25] Comparison of system COP in three regions 78
[Figure 3.26] On-site arrangement of the scenario 1 81
[Figure 3.27] On-site arrangement of the scenario 2 85
[Figure 3.28] On-site arrangement of the scenario 3 88
[Figure 3.29] Internal view of the HPDC-U system 90
[Figure 3.30] Indoor environment changes in the FT1 case 93
[Figure 3.31] Indoor environment changes in the FT2 case 93
[Figure 3.32] Indoor comfort comparison between the FT1 and FT2 94
[Figure 3.33] Indoor air and supply air state changes during FT3 94
[Figure 3.34] Indoor air state under different humidity control types 96
[Figure 3.35] Improvement effect of indoor thermal environment based on operation mode control 97
[Figure 3.36] Power and electricity consumption of FT1 and FT2 99
[Figure 3.37] Comparison of electricity use between the HPDC system and EHP under high load conditions 100
[Figure 3.38] System air state changes during the experiment period 102
[Figure 3.39] System sensible and latent cooling capacity for each LP case. 103
[Figure 3.40] Regeneration temperature obtained by the condensing heat 104
[Figure 3.41] Indoor temperature and humidity change after adopting OMAS control 107
[Figure 3.42] Process inlet and outlet air state change during OMAS control 107
[Figure 3.43] Variation in latent heat removed by the evaporator and desiccant wheel 108
[Figure 4.1] Structure and layout of the WLDC system used in experiments 112
[Figure 4.2] Air treatment process of the WLDC system in a psychrometric chart 113
[Figure 4.3] Control logic of WLDC system 114
[Figure 4.4] Internal view of the secondary side of the WLDC system used in the validation experiment 116
[Figure 4.5] Structure and layout of the primary side of the WLDC system 117
[Figure 4.6] Experiment site setup of the secondary side of the WLDC system 118
[Figure 4.7] Experiment site setup of the primary side of the WLDC system 120
[Figure 4.8] Changes in the indoor air state during the WLDC system operation 123
[Figure 4.9] Air state in the dehumidification side of the WLDC system 124
[Figure 4.10] Thermal energy changes occurring in each component 125
[Figure 4.11] Supply water temperature of cooling coils 125
[Figure 4.12] Comparison between the primary energy 127
[Figure 4.13] Diagram of the high temperature cooling terminal system structure 130
[Figure 4.14] Layout of the validation experiment 137
[Figure 4.15] Surface temperature of radiant panels in experiment and simulation 138
[Figure 4.16] 3-D view of the simulation building models 139
[Figure 4.17] Indoor load results during summer season 141
[Figure 4.18] Fraction schedule of different indoor heat source 141
[Figure 4.19] High temperature cooling terminal system decoupling operation 143
[Figure 4.20] High temperature cooling terminal system decoupling operation with cool-down time control 145
[Figure 4.21] Reset control added in the decoupling operation control strategy 147
[Figure 4.22] Indoor air temperature changes under the decoupling operation control scheme without cool-down time control 149
[Figure 4.23] Indoor air relative humidity changes under the decoupling operation control scheme 150
[Figure 4.24] Indoor air temperature comparison of decoupling operation schemes with different temperature control schemes 151
[Figure 4.25] Control action of different reset methods 152
[Figure 4.26] Distribution of RCH values[이미지참조] 156
[Figure 4.27] Variation of RCH value distribution under different indoor design relative humidity[이미지참조] 157
[Figure 4.28] Variation of RCH value distribution under different indoor design temperature[이미지참조] 158
[Figure 4.29] Average RCH value of the system under different indoor design temperature and humidity conditions[이미지참조] 159
[Figure 4.30] Principal diagram of DDPG algorithm 162
[Figure 4.31] Determining process of control task. 163
[Figure 4.32] Process of centralized training and decentralized execution 163
[Figure 4.33] Pseudo-code for MADDPG 166
[Figure 4.34] Co-simulation framework for MADDPG simulation. 167
[Figure 4.35] Total and object reward changes during training phase 169
[Figure 4.36] Indoor design temperature and relative humidity changes under MADDPG algorithm control 170
[Figure 4.37] Indoor PMV change when using MADDPG algorithm control and rule-based control 171
[Figure 4.38] Comparison of system RCH value under different control methods[이미지참조] 172
[Figure 5.1] Heat source system configuration of the proposed THIC system 185
[Figure 5.2] Layout of the conventional THIC system and its heat source system 186
[Figure 5.3] Diagram of the internal structure and operation process of the central heat pump 186
[Figure 5.4] Diagram of the heat recovery water tank model used in the THIC system 187
[Figure 5.5] Cooling capacity of the heat source system in different THIC systems 189
[Figure 5.6] Comparison of cooling loads of cooling components in different THIC systems 190
[Figure 5.7] Percentage of component cooling load to the overall system cooling load 190
[Figure 5.8] COP of the heat source system in different THIC systems 191
[Figure 5.9] Part load ratio of chillers in different heat source systems 191
[Figure 5.10] Changes in the condenser inlet water temperature of chillers in different heat source systems 192
[Figure 5.11] Energy saving effect of the proposed THIC system 193
[Figure 5.12] Ratio of reduction in COP and cooling load of chiller in the proposed THIC system 193
[Figure 5.13] Electricity saving of the proposed THIC system compared to the conventional THIC system 194
[Figure 5.14] Heating energy used in different THIC systems 194
[Figure 5.15] Heat recovery outlet and tank water temperature changes 196
[Figure 5.16] Relationship between the heat recovery tank water temperature and outdoor air humidity ratio 196
[Figure 5.17] Recovery percentage of condensing heat under different outdoor air conditions 198
[Figure 5.18] Cooling capacity of each component as a percentage of total system cooling capacity 199
[Figure 5.19] Agent design in the multi-zone applications 201
[Figure 5.20] Rewards variation in training phases. 202
[Figure 5.21] Chiller average electricity use in training phase 203
[Figure 5.22] Average PMV of all office areas during training 203
[Figure 5.23] Electricity consumption comparison between different control methods 205
[Figure 5.24] Chilled water temperature under different system control methods 206
[Figure 5.25] MRC and chiller operation time under different system control methods 206
[Figure 5.26] Indoor PMV changes during test period under two control methods. (a) MADDPG algorithm control; (b) rule-based control. 207