Chapter 1  Introduction
  1.1  Motivations and significance
  1.2  Feature modeling methods for industrial processes
    1.2.1  Single modal feature modeling approach
    1.2.2  Multimodal feature modeling approach
  1.3  Operating condition evaluation methods for industrial processes
    1.3.1  Conventional operating condition evaluation methods
    1.3.2  Operating condition evaluation methods based on feature modeling
Chapter 2  Hot Strip Rolling Mill Process
  2.1  Introduction
  2.2  Description of HSRM process
    2.2.1  Process principle of HSRM process
    2.2.2  Process flow of HSRM process
  2.3  Data characteristics of HSRM process
  2.4  Multi-condition characteristics of HSRM process
    2.4.1  The HSRM process with multiple operating conditions
    2.4.2  Multiple operating conditions modeling strategy
  2.5  Control of FR process and analysis of operating conditions
    2.5.1  FR process control analysis
    2.5.2  Analysis of operating conditions
  2.6  Conclusions
Chapter 3  Rolling Process Fault Injection System
  3.1  Introduction
  3.2  Rolling process common faults
  3.3  Rolling process fault injection model
    3.3.1  Bounce equation
    3.3.2  Forward sliding model
    3.3.3  Temperature drop model
    3.3.4  Rolling force model
    3.3.5  Bend roll force model
  3.4  Rolling process fault injection system and its realization
    3.4.1  Fault injection principle for HSRM processes
    3.4.2  Fault injection system visualization design
  3.5  Data presentation for several types of faults
  3.6  Conclusions
Chapter 4  A Unified Common-Specific Feature Modeling Framework
  4.1  Introduction
  4.2  Common and specific features of the hot rolling process
  4.3  A framework for modeling common-specific features
    4.3.1  Common feature extraction method based on relevant process variables
    4.3.2  Common feature extraction method based on public score matrix
    4.3.3  Common feature extraction method based on common weight matrix
    4.3.4  Common feature extraction based on common basis vectors
    4.3.5  ICA-based specific feature extraction for multimodal processes
  4.4  Feature modeling and condition evaluation
    4.4.1  Evaluation of operating conditions based on the T2 statistic
    4.4.2  Evaluation of operating conditions based on the KL divergence statistic
  4.5  Conclusions
Chapter 5  A Feature Modeling Approach Based on Common-Specific PLS
  5.1  Introduction
  5.2  Common and specific subspaces
  5.3  The CnS-PCA-based PM method
    5.3.1  PCA-based common subspace
    5.3.2  Further modeling in the specific subspace
    5.3.3  Process monitoring based on CnS-PCA model
  5.4  The CnS-PLS-based method
    5.4.1  Quality-relevant common and specific subspaces
    5.4.2  CnS-PLS-based process monitoring method
  5.5  Conclusions
Chapter 6  A Feature Modeling Approach Based on Common-Specific Subspaces
  6.1  Introduction
  6.2  Description of common and specific subspaces
  6.3  The extraction of common and specific features
    6.3.1  Common subspace
    6.3.2  Quality-relevant common and specific subspaces
  6.4  Common subspace migration
  6.5  The modeling results of common-specific subspaces
  6.6  KL-based monitoring metric based on SPD matrices
  6.7  Conclusions
Chapter 7  Feature Modeling Method Based on Tensor Decomposition
  7.1  Introduction
  7.2  Tensor decomposition methods
    7.2.1  Tensor overview
    7.2.2  Tensor decomposition
  7.3  Tensor decomposition of common features extraction
    7.3.1  Construction of CSu using canonical polyadic decomposition (CPD)
    7.3.2  Calculation based on ALS
  7.4  Tensor decomposition personality characteristics modeling
  7.5  Summary and comparison of tensor decomposition methods
  7.6  Conclusions
Chapter 8  Feature Modeling Method Based on Common-Specific DBN
  8.1  Introduction
  8.2  Description of the DBN
    8.2.1  Restricted boltzmann machine(RBM)
    8.2.2  DBN training process
    8.2.3  Fault diagnosis method based on DBN
  8.3  Feature extraction based on CnS-DBN
    8.3.1  Feature extraction network architecture
    8.3.2  DBN Pre-training
    8.3.3  Common and specific feature extraction
  8.4  Method characteristics and comparative analysis
  8.5  Conclusions
Chapter 9  Operating State Evaluation Based on Linear Feature Modeling Methods
  9.1  Introduction
  9.2  Evaluation metrics definition
  9.3  Evaluation results of CnS-PCA-based method
    9.3.1  Three fault cases
    9.3.2  CnS-PCA-based monitoring results
    9.3.3  CnS-PLS-based monitoring results
  9.4  Common-specific subspace-based operating state evaluation results
    9.4.1  Faults affecting the common subspace
    9.4.2  Faults affecting the specific subspace
    9.4.3  Fault affecting both the common and specific subspace
    9.4.4  Comparison and discussion
  9.5  Evaluation results based on tensor decomposition
    9.5.1  Tensor decomposition-based common subspace method
    9.5.2  The structural fault monitoring results
    9.5.3  The non-structural fault monitoring results
    9.5.4  Comparisons of the monitoring performance using statistical results
  9.6  Conclusions
Chapter 10  Operating State Evaluation Based on Common-Specific DBN
  10.1  Introduction
  10.2  Evaluation metrics definition
  10.3  Abnormal operating condition model training
    10.3.1  Parameters selection
    10.3.2  Model training process
  10.4  Abnormal operating condition evaluation results
    10.4.1  Operation condition evaluation with known fault information
    10.4.2  Operation condition evaluation with known partial fault information
  10.5  Conclusions
Chapter 11  Operating Condition Evaluation Prototype System of HSRM Process
  11.1  Introduction
  11.2  Prototype system hardware framework
  11.3  Prototype system software framework
    11.3.1  Database software
    11.3.2  Cloud computing server software configuration
    11.3.3  Data playback platform software configuration
    11.3.4  Docker container deployment and application
  11.4  Prototype system data playback and presentation
    11.4.1  Data playback system communication principles
    11.4.2  Fault data interacts with real-time databases
    11.4.3  Data playback function validation
  11.5  Conclusions
Chapter 12  Real-Time Evaluation of Operating Conditions
  12.1  Introduction
  12.2  The real-time evaluation architecture using cloud-edge-end collaboration
  12.3  Real-time data alignment and pre-processing for FMP
    12.3.1  Multi-stand data alignment method
    12.3.2  Real-time data alignment and preprocessing
  12.4  Edge real-time evaluation case presentation
  12.5  Conclusions
References