Cover
Title Page
Preface
Biography
1 Introduction to Structural Health Monitoring
1. 1.1 Advances in Structural Health Monitoring Technology
2. 1.2 Structural Health Monitoring System and Strategy
3. 1.3 Potential Benefits of SHM in Civil Engineering
4. 1.4 Challenges and Further Work of SHM
5. 1.5 Concluding Remarks
6. References
2 Sensors and Sensing Technology for Structural Monitoring
1. 2.1 Introduction
2. 2.2 Sensor Types
3. 2.3 Sensor Measurements in Structural Monitoring
4. 2.4 Fibre Optic Sensors
5. 2.5 Wireless Sensors
6. 2.6 Optimum Sensor Selection and Placement
7. 2.7 Case Study
8. 2.8 Concluding Remarks
9. References
3 Data Acquisition, Transmission and Management
1. 3.1 Introduction
2. 3.2 Data Acquisition Systems
3. 3.3 Data Transmission Systems
4. 3.4 Data Processing Systems
5. 3.5 Data Management Systems
6. 3.6 Case Study
7. 3.7 Concluding Remarks
8. References
4 Structural Damage Identification Techniques
1. 4.1 Introduction
2. 4.2 Damage in Structures
3. 4.3 Non‐Destructive Testing Techniques
4. 4.4 Comparison of NDT and SHM
5. 4.5 Signal Processing for Damage Detection
6. 4.6 Data‐Based Versus Model‐Based Techniques
7. 4.7 Development of Vibration‐Based Methods
8. 4.8 Concluding Remarks
9. References
5 Modal Analysis of Civil Engineering Structures
1. 5.1 Introduction
2. 5.2 Basic Equations for Structural Dynamics
3. 5.3 Input‐Output Modal Identification
4. 5.4 Output‐Only Modal Identification
5. 5.5 Correlation Between Test and Calculated Results
6. 5.6 Mode Shape Expansion and Model Reduction
7. 5.7 Case Study
8. 5.8 Concluding Remarks
9. References
6 Finite Element Model Updating
1. 6.1 Introduction
2. 6.2 Finite Element Modelling
3. 6.3 Structural Parameters for Model Updating
4. 6.4 Sensitivity Based Methods
5. 6.5 Dynamic Perturbation Method
6. 6.6 Use of Dynamic Perturbation Method for Model Updating
7. 6.7 Case Study
8. 6.8 Concluding Remarks
9. References
7 Vibration‐Based Damage Identification Methods
1. 7.1 Introduction
2. 7.2 Structural Modelling for Damage Identification
3. 7.3 Methods Using Change of Modal Parameters
4. 7.4 Methods Using Change of Structural Parameters
5. 7.5 Pattern Recognition Methods
6. 7.6 Neural Network Techniques
7. 7.7 Concluding Remarks
8. References
8 Model‐Based Damage Assessment Methods
1. 8.1 Introduction
2. 8.2 Characterisation of Damage in Structures
3. 8.3 Matrix Update Methods
4. 8.4 Sensitivity Based Methods
5. 8.5 Damage Assessment Using Dynamic Perturbation Method
6. 8.6 Numerical Examples
7. 8.7 Potential Problems in Vibration‐Based Damage Identification
8. 8.8 Concluding Remarks
9. References
9 Monitoring Based Reliability Analysis and Damage Prognosis
1. 9.1 Introduction
2. 9.2 Usage Monitoring
3. 9.3 Probabilistic Deterioration Modelling
4. 9.4 Lifetime Distribution Analysis
5. 9.5 Structural Reliability Analysis
6. 9.6 Optimum Maintenance Strategy
7. 9.7 Case Study
8. 9.8 Concluding Remarks
9. References
10 Applications of SHM Strategies to Large Civil Structures
1. 10.1 Introduction
2. 10.2 SHM System and Damage Identification of a Cable‐Stayed Bridge
3. 10.3 In‐Construction Monitoring of a High‐Rise Building
4. 10.4 Monitoring of Tunnel Construction Using FBG Sensors
5. 10.5 Safety Monitoring of Rail Using Acoustic Emission
6. 10.6 Structural Integrity Monitoring of Water Mains
7. 10.7 Concluding Remarks
8. References
Index
End User License Agreement

List of Tables

Chapter 02
1. Table 2.1 Typical sensing methods for structural health monitoring.
2. Table 2.2 Examples of sensors used for monitoring of structural responses and operation loads.
3. Table 2.3 Examples of sensors used for monitoring of environmental factors.
4. Table 2.4 Typical physical quantities of structural responses for monitoring.
5. Table 2.5 Typical physical and chemical quantities of environmental factors for monitoring.
6. Table 2.6 Typical physical quantities of operational loads for monitoring.
7. Table 2.7 Typical physical quantities of structural characteristics for monitoring.
Chapter 04
1. Table 4.1 Comparison of data‐based and model‐based damage assessment techniques.
Chapter 05
1. Table 5.1 Comparison of identified natural frequencies and damping ratios between the three excitation types (after Peeters and Ventura 2003).
2. Table 5.2 Identified natural frequencies (Hz) of the first eight modes under various wind conditions using the SSI‐DATA method and the SSI‐COV method.
3. Table 5.3 Modal data identified by output‐only identification methods from ambient vibration data.
Chapter 06
1. Table 6.1 Simulated, measured and updated frequencies of the cantilever beam without weighting on initial parameter estimates.
2. Table 6.2 Simulated, measured and updated frequencies of the cantilever beam with weighting on initial parameter estimates.
3. Table 6.3 Updated natural frequencies at different iteration numbers for model updating using only eight ‘tested’ frequencies.
4. Table 6.4 Updated natural frequencies at different iteration numbers for model updating using three ‘tested’ incomplete modes.
5. Table 6.5 Both stiffness and mass factors adjusted simultaneously using different number of ‘tested’ incomplete modes.
6. Table 6.6 Updated natural frequencies of the FE model using eight simulated incomplete modes without noise and with 7% noise level.
7. Table 6.7 Comparison of updated natural frequencies of the FE model using six real experimental incomplete modes by different model updating methods.
8. Table 6.8 Updated modal properties of the reduced FE model using 10 measured frequencies excluding two torsion modes.
9. Table 6.9 Updated modal properties of the reduced FE model using nine measured incomplete modes.
Chapter 07
1. Table 7.1 First five frequencies (Hz) of the cantilever beam for different model problems.
2. Table 7.2 Comparison of measured and computed natural frequencies (Hz) of Tsing Ma Bridge.
3. Table 7.3 Simulated damage cases of Tsing Ma Bridge.
4. Table 7.4 Results of damage location identification using the COMAC method.
5. Table 7.5 Results of damage location identification using the ECOMAC method.
6. Table 7.6 Results of damage location identification using the mode shape curvature (MSC) method.
7. Table 7.7 Results of damage location identification using the flexibility matrix method.
8. Table 7.8 Results of damage location identification using the strain energy based damage index method.
9. Table 7.9 Comparison of different damage identification methods applied to Tsing Ma Bridge.
10. Table 7.10 Results of simulated damaged member identification in deck segment No.3 using artificial neural network (ANN) with back‐propagation (BP).
11. Table 7.11 Damage location identification results for probabilistic neural network (PNN).
Chapter 08
1. Table 8.1 Values of damage indicators for cases with loss of bending stiffness capacity at critical locations.
2. Table 8.2 First five natural frequencies (Hz) for original and damaged structure.
3. Table 8.3 First four natural frequencies (Hz) and original mode shapes.
4. Table 8.4 Geometric and material properties of the structural members of the framed building structure.
5. Table 8.5 Simulated damage scenarios for the framed building structure.
6. Table 8.6 Correlated noise free frequencies (Hz) of the original and damaged structure.
7. Table 8.7 Simulated damage scenarios for the gravity dam structure.
Chapter 09
1. Table 9.1 List of instrumented structural components and monitoring parameters.
Chapter 10
1. Table 10.1 Recorded ambient acceleration data samples under different wind conditions.
2. Table 10.2 Identified frequencies (Hz) of first eight modes under various wind speed conditions.
3. Table 10.3 Simulated damage cases for damage detection studies of Ting Kau Bridge.
4. Table 10.4 Damage location identification using mode shape curvature index Z‐values.
5. Table 10.5 Damage‐caused frequency change and damage detectability.
6. Table 10.6 Summary of correct damage identification using probabilistic neural network (PNN) technique.
7. Table 10.7 Comparison between identified and predicted natural frequencies (Hz).

List of Illustrations

Chapter 01
1. Figure 1.1 System architecture and operation diagram of a SHM system for a bridge.
2. Figure 1.2 Structural health monitoring strategies for civil engineering structures (after Frangopol and Messervey 2009).
3. Figure 1.3 Integrated framework for health monitoring and evaluation of civil structures.
4. Figure 1.4 SHM strategy for health evaluation and maintenance planning of civil structures.
Chapter 02
1. Figure 2.1 Schematic diagram of a typical sensory system, converting measured quantities into analogue or digital signals.
2. Figure 2.2 Accelerometers used in vibration monitoring of civil engineering structures.
3. Figure 2.3 Global positioning system (GPS, Leica Geosystems Model 1230) used in construction monitoring of Canton Tower.
4. Figure 2.4 Vibrating wire strain gauges (Geokon Model GK4200) installed at the inner tube of Canton Tower.
5. Figure 2.5 Anemometer (R.M. Young Model 05103L) installed on Canton Tower.
6. Figure 2.6 Deployment of weigh‐in‐motion (WIM) sensors on Tsing Ma Bridge.
7. Figure 2.7 Instrumentation layout of sensory systems for structural monitoring in Tsing Ma Bridge.
8. Figure 2.8 Modular architecture and input/output block diagrams of the monitoring system of Tsing Ma Bridge.
9. Figure 2.9 Setup of SOFO interferometric sensor system.
10. Figure 2.10 Functional principle of Fabry‐Pérot sensors.
11. Figure 2.11 Schematic illustration of a fibre Bragg grating (FBG) sensor.
12. Figure 2.12 Wireless sensors for structural monitoring.
13. Figure 2.13 Functional subsystem and block diagrams of wireless sensors (after Lynch and Loh 2006).
14. Figure 2.14 Photo of Canton Tower.
15. Figure 2.15 Vision inspection system of Canton Tower.
16. Figure 2.16 Deployment of sensors and data acquisition substations on Canton Tower for in‐service monitoring.
17. Figure 2.17 Fibre Bragg gratings (FBG) sensing system for Canton Tower.
Chapter 03
1. Figure 3.1 Schematic layout of a typical data acquisition unit and its associated equipment.
2. Figure 3.2 Comparison of wired and wireless configurations of a structural monitoring system (after Straser et al. 1998).
3. Figure 3.3 Schematic layout of data acquisition and transmission system.
4. Figure 3.4 Block diagram of traffic data analysis in DPCS‐2, combined digital video converter (DVC) and dynamic weigh‐in‐motion (DWIM) data.
5. Figure 3.5 Architectural layout of structural health data management system (SHDMS) and its interfaces.
6. Figure 3.6 Integrated in‐construction and in‐service SHM system for Canton Tower.
7. Figure 3.7 Integration of SHM and vibration control.
8. Figure 3.8 Wireless data transmission.
Chapter 04
1. Figure 4.1 Traditional NDT techniques depending on wave types (after Ettouney and Alampalli 2012).
2. Figure 4.2 Equipment and applications of the guided wave technique.
3. Figure 4.3 Schematic diagrams of a laser Doppler vibrometer system.
4. Figure 4.4 Schematic diagrams of configuration of global positioning system (GPS) components.
5. Figure 4.5 Data‐based and model‐based damage assessment techniques in structural health monitoring strategies (after Farrar and Lieven 2007).
Chapter 05
1. Figure 5.1 Impact hammer and accelerometers for vibration testing.
2. Figure 5.2 A space steel frame structure used for laboratory testing with test equipment.
3. Figure 5.3 Laboratory vibration test results for the space steel frame structure.
4. Figure 5.4 Experimental modes from laboratory vibration test for the space frame structure.
5. Figure 5.5 Fundamental frequency versus hourly‐average temperature of the Ting Kau cable‐stayed bridge.
6. Figure 5.6 Relative frequencies extracted by various modal identification methods from the ambient vibration data (after Peeters and Ventura 2003).
7. Figure 5.7 Relative damping ratios extracted by various modal identification methods from the drop weight data (after Peeters and Ventura 2003).
8. Figure 5.8 Deployment of accelerometers on Ting Kau Bridge.
9. Figure 5.9 Stabilisation diagrams for the SSI‐DATA and the SSI‐COV output‐only identification methods for Ting Kau Bridge.
10. Figure 5.10 Identified mode shapes of first eight modes for Ting Kau Bridge.
11. Figure 5.11 Plane frame structure with possible sensor locations marked with a circle.
12. Figure 5.12 Mean cumulative errors in mode shape expansion as a function of noise level in DOF’s readings for the frame structure.
13. Figure 5.13 Overall orthogonality errors in mode shape expansion as a function of noise level in DOF readings for the frame structure.
14. Figure 5.14 Canton Tower: (a) full finite element model, (b) positions of installed accelerometers, (c) reduced‐order finite element (FE) model.
15. Figure 5.15 Averaged normalised power spectral densities (ANPADs) of the measured acceleration data used for the peak picking (PP) method.
16. Figure 5.16 Stabilisation diagram of measured acceleration data used for the stochastic subspace identification (SSI) method.
17. Figure 5.17 Expanded first and fourth bending modes, compared with the modal readings identified from ambient vibration data and the eigenvectors of the reduced finite element model.
Chapter 06
1. Figure 6.1 A cantilever beam structure for model updating.
2. Figure 6.2 Finite element model of a plane frame structure for model updating.
3. Figure 6.3 Finite element model of the laboratory tested space steel frame structure.
4. Figure 6.4 Comparison of updated stiffness parameters of the FE model and simulated exact stiffness parameters at integration point level for beams and columns and at element level for braces, information on different number of noise‐free incomplete modes used.
5. Figure 6.5 Updated stiffness parameters of the reduced FE model using nine measured incomplete modes.
Chapter 07
1. Figure 7.1 Possible global structural damage indices for civil engineering structures (after Catbas and Aktan 2002).
2. Figure 7.2 A cantilever beam used for different modelling problems.
3. Figure 7.3 Elevation of Tsing Ma Bridge.
4. Figure 7.4 Percentage difference of natural frequencies (PDNF) for simulated damage Cases 1.1 and 1.2.
5. Figure 7.5 Mode shape curvature (MSC) values for damage Case 2.1 using 3 damaged modes.
6. Figure 7.6 Flexibility coefficient index values at nodes on deck for damage Case 1.2.
7. Figure 7.7 Values of the strain energy based damage index β_j for damage Case 2.1.
8. Figure 7.8 Simulated ‘normal’ fluctuation of tension force of the cable.
9. Figure 7.9 Novelty index evaluated on training and testing data with 5% reduction in cable tension.
10. Figure 7.10 Back‐propagation (BP) neural network with three layers and nodes.
11. Figure 7.11 Architecture of a three‐layer probabilistic neural network (PNN) for damage identification.
12. Figure 7.12 Longitudinal frame member numbering in deck segment No.3 and the selected mode shape components marked with arrows.
Chapter 08
1. Figure 8.1 Statically indeterminate plane truss structure for damage assessment.
2. Figure 8.2 Inverse predictions for the hypothetical damage scenario of the plane truss using the residual force vector method.
3. Figure 8.3 Simply supported grid structure (plane view).
4. Figure 8.4 Inverse damage identification from the sensitivity based method using natural frequencies.
5. Figure 8.5 Inverse damage identification using the dynamic perturbation method from 5 damaged frequencies.
6. Figure 8.6 One‐bay six‐bar plane truss structure.
7. Figure 8.7 Inverse damage identification using the dynamic perturbation method from incomplete modes.
8. Figure 8.8 Finite element model of the framed building structure with element numbering and sensor locations (marked with ⊗).
9. Figure 8.9 Singular values, ordinary solution coefficients and regularised solution coefficients for inverse damage predictions.
10. Figure 8.10 L‐curves for Tikhonov regularisation with noise levels ranging from 1% to 5% in measured DOF readings.
11. Figure 8.11 Identification of damage scenario I (no stiffness in one brace in first storey, i.e. element no. 24), 10 incomplete noisy damaged modes used.
12. Figure 8.12 Identification of damage scenario II (no stiffness in braces in first storey, i.e. element nos. 22–29), 7 incomplete noisy damaged modes used.
13. Figure 8.13 Gravity dam structure with measured nodes marked with both ▪ and ●.
14. Figure 8.14 Inverse damage predictions using the dynamic perturbation method for the gravity dam structure.
Chapter 09
1. Figure 9.1 Framework for reliability analysis using an SHM strategy (after Catbas et al. 2008).
2. Figure 9.2 Monitoring of wind velocities, directions and occurrence at tower‐top of Tsing Ma Bridge.
3. Figure 9.3 Axle load distribution and spectrum from the weigh‐in‐motion data.
4. Figure 9.4 Gross vehicle weight distribution and spectrum from the weigh‐in‐motion data.
5. Figure 9.5 Fatigue crack growth curves simulated by the gamma process model and predicted by the Paris–Erdogan law.
6. Figure 9.6 Probability distribution of time to failure using the gamma process model under different critical thresholds of fatigue crack length (a_cr).
7. Figure 9.7 Lifetime distribution (CDF) as a function of concrete crack width.
8. Figure 9.8 Symptom reliability (survival function) by Bayesian updating due to inspection and repair.
9. Figure 9.9 Probability density functions (PDFs) of resistance effect R and load effect S.
10. Figure 9.10 Schematic representation of load processes and resistance deterioration (after Melchers 1999).
11. Figure 9.11 Structural lifetime T^* associated with a target probability of failure .
12. Figure 9.12 Deck cross‐section CH24662.5 and strain gauge SPTLS16 at Detail H of Tsing Ma Bridge.
13. Figure 9.13 Monitoring data of dynamic stresses collected from strain gauge SPTLS16.
14. Figure 9.14 Probability of failure and reliability index as a function of fatigue life.
15. Figure 9.15 Pareto optimum solution sets with single and double interventions for inspection and maintenance.
16. Figure 9.16 Dynamic strain gauges in longitudinal stiffening truss for highway traffic monitoring.
17. Figure 9.17 Estimates of railway loading based on strain results in waybeams.
18. Figure 9.18 Comparison of measured and designed suspender forces.
19. Figure 9.19 Stress history in top chord of outer longitudinal truss (North).
20. Figure 9.20 Fatigue life estimation for top chord of outer longitudinal truss (North).
Chapter 10
1. Figure 10.1 Three‐tower cable‐stayed Ting Kau Bridge.
2. Figure 10.2 Modular architecture of structural monitoring systems for Ting Kau Bridge.
3. Figure 10.3 Sensory system of SHM strategy for Ting Kau Bridge.
4. Figure 10.4 Three‐dimensional finite element model of Ting Kau Bridge.
5. Figure 10.5 Framework for finite element model updating by measurement data.
6. Figure 10.6 New Headquarters of Shenzhen Stock Exchange, a high‐rise building with a floating platform.
7. Figure 10.7 Layout of temporary braces of the high‐rise building.
8. Figure 10.8 Growth of stresses measured in the process of dismantling temporary supports.
9. Figure 10.9 Comparison of measured and computed stress increments during unloading process.
10. Figure 10.10 Wireless monitoring system for the high‐rise building.
11. Figure 10.11 Distribution of accelerometers on the floating platform.
12. Figure 10.12 Vertical dynamic response of floating platform at sensor location P9.
13. Figure 10.13 River‐crossing metro tunnel during freezing construction.
14. Figure 10.14 Freezing monitoring tube accommodating a FBG sensor.
15. Figure 10.15 Deployment locations of FBG sensors in the frozen soils.
16. Figure 10.16 Measured temperature time histories during freezing construction.
17. Figure 10.17 Deployment of FBG based liquid‐level sensors on subgrade surface of the railway line.
18. Figure 10.18 Measured Bragg wavelengths by liquid‐level sensors for estimating subgrade settlement.
19. Figure 10.19 Piezoelectric (PZT) based acoustic emission sensing system for damage detection of rail turnouts.
20. Figure 10.20 Undamaged rail turnouts and arrangement of PZT transducers on railway switch area.
21. Figure 10.21 Signals collected from PZT transducer 1 for undamaged rail turnout.
22. Figure 10.22 Damaged rail turnout and passage train.
23. Figure 10.23 Signals collected from PZT transducer 1 for damaged rail turnout.
24. Figure 10.24 Comparisons of power spectrum density (PSD) from PZT transducer 1 for undamaged and damaged rail turnout.
25. Figure 10.25 Schematic of FBG sensory system for structural integrity monitoring of water mains.
26. Figure 10.26 FBG sensors for monitoring water mains.
27. Figure 10.27 Distribution of strain, temperature and pressure sensors.
28. Figure 10.28 Topology of the sensory system for structural monitoring of water mains.
29. Figure 10.29 Correlation of strain and temperature variation at measurement point A with water flow.
30. Figure 10.30 Correlation of strain and temperature variation at measurement point A with water cutoff.
31. Figure 10.31 Strain and temperature changes at cross‐section #35 with water cutoff.
32. Figure 10.32 Water pressures at the middle of the GRP pipeline with water cutoff.
33. Figure 10.33 Normalised cross‐correlation between each pair of temperature and strain sensors at cross‐section #35, where blue (solid) line for point A; green (dashed) line for point B; red (dotted) line for point C.

Preface

Civil engineering structures such as bridges and buildings are typically large and are built with uncertainties. Their behaviour during the construction phase should be monitored to control the quality and safety of the construction processes. After civil structures have been constructed, the construction materials are subjected to degradation over time, leading to a decrease in structural capacity and serviceability. Monitoring during the service phase offers useful information on structural performance under gradual material degradation and expected loads, and also records the structural responses of unexpected sudden overloading. Data collected from real time monitoring can then be used for damage assessment and health evaluation of the civil engineering structures in service. The continuously measured data from the monitoring system can provide the basis for predicting future performance and determining optimum maintenance strategy for the existing structures.

Structural health monitoring (SHM) is a process of in‐service damage identification and health evaluation for an engineering structure through an automated monitoring system. SHM uses sensing systems and necessary hardware and software facilities to monitor structural responses and operational conditions of the structure. A typical SHM strategy comprises several key components, including sensors, data acquisition, data transmission, data processing, data management, health evaluation and decision making. Sensing technology and the signal interpretation algorithms are two critical factors in developing successful SHM strategies for large civil engineering structures. Damage assessment methods using vibration measurements such as modal parameters show promise for the health evaluation of the civil structures.

The development of a structural health monitoring strategy requires a multidisciplinary approach involving many fields, such as sensors and sensor networks, signal processing, modal testing, numerical modelling, probabilistic analysis, damage diagnosis and damage prognosis. Each of these topics is a discipline‐specific subject by itself, and is equally important in developing effective SHM strategies. Sensing systems are critical for accurate data acquisition and transmission, and the acquired data is used for signal processing to extract key features sensitive to local damage. Modal testing and analysis is adopted to identify modal parameters from vibration measurements, and the obtained modal data can be used for model updating and damage assessment. Probabilistic approaches are needed for numerical modelling to account for uncertainties, and provide an essential framework for reliability analysis and damage prognosis. The objective of this book is to integrate these topics with the specific focus on developing SHM strategies for large civil engineering structures.

This book aims to explain the principles of the SHM strategy, and so it covers all aspects of sensing system, data processing and analysis, damage assessment and decision making for structural monitoring and health evaluation of large civil engineering structures. The book consists of four major parts. First, sensors and sensing technology and data transmission systems are introduced for monitoring of civil structures. From the data measured from the monitoring system on the structure, modal analysis techniques are presented to extract modal parameters, which are used to update and validate the associated finite element numerical model. Then, various methods are provided for identifying the existence, location and extent of damage in civil structures using the measured data and their derivatives. Finally, from the continuously monitored data, probabilistic approaches are utilised for deterioration modelling and reliability analysis, giving the basis for decision making. The techniques for the SHM strategy are well explained in a number of examples and are also demonstrated in many real case studies.

This book can be used as the textbook for a graduate level course on structural health monitoring with emphasis on civil engineering structures. Also, the book can be used as a guide for the practising engineers who want to apply SHM techniques in practice. The book is written with an assumption that the reader has a basic engineering background and needs knowledge of little more than undergraduate level mathematics. Furthermore, the book is an invaluable reference for those undertaking research in the areas of structural monitoring and health evaluation of civil engineering structures.

In this book, several real case studies on health monitoring of civil engineering structures, such as Tsing Ma Bridge, Ting Kau Bridge and Canton Tower, are generously contributed by Professor Yi‐Qing Ni, The Hong Kong Polytechnic University. These practical applications cover various areas in structural health monitoring technology including sensors and sensing networks, data transmission and processing systems, structural damage identification techniques and usage monitoring systems, which are used as examples in several chapters, such as Chapter 2, Chapter 3, Chapter 7 and Chapter 10. The book would not be complete without these practical examples, thus deepest gratitude must go to Professor Yi‐Qing Ni and his colleagues, in particular Professor J.M. Ko, Dr. K.Y. Wong and Dr. X.W. Ye.

Finally, the author is indebted to many people for their direct and indirect assistance in the preparation of this book. The author would like to thank the former and current colleagues, research fellows and PhD students for their support and useful works, particularly Dr. T.L. Huang, Dr. T.S. Maung, Dr. J. Nepal and Mr. C. Zhang. The author deeply thanks his family for their continuous patience and understanding; especially Chengheng Xiao, Helen, Alice and Xuezhang have been the constant supporters.

November 2017

Hua‐Peng Chen, in London