Model-Based Fault Diagnosis Techniques Design Schemes, Algorithms and Tools

Guaranteeing a high system performance over a wide operating range is an important issue surrounding the design of automatic control systems with successively increasing complexity. As a key technology in the search for a solution, advanced fault detection and identification (FDI) is receiving considerable attention. This book introduces basic model-based FDI schemes, advanced analysis and design algorithms, and mathematical and control-theoretic tools.
This second edition of Model-Based Fault Diagnosis Techniques contains:
• new material on fault isolation and identification and alarm management;
• extended and revised treatment of systematic threshold determination for systems with both deterministic unknown inputs and stochastic noises;
• addition of the continuously-stirred tank heater as a representative process-industrial benchmark; and
• enhanced discussion of residual evaluation which now deals with stochastic processes.
Model-based Fault Diagnosis Techniques will interest academic researchers working in fault identification and diagnosis and as a text it is suitable for graduate students in a formal university-based course or as a self-study aid for practising engineers working with automatic control or mechatronic systems from backgrounds as diverse as chemical process and power engineering.

Basic Ideas, Major Issues, and Tools in the Observer-Based FDI Framework.- Modelling of Technical Systems.- Structural Fault Detectability, Isolability and Identifiability.- Basic Residual Generation Methods.- Perfect Unknown Input Decoupling.- Residual Generation with Enhanced Robustness against Unknown Inputs.- Residual Generation with Enhanced Robustness against Model Uncertainties.- Norm-Based Residual Evaluation and Threshold Computation.- Statistical-Methods-Based Residual Evaluation and Threshold Setting.- Integration of Norm-Based and Statistical Methods.- Integrated Design of Fault Detection Systems.- Fault Isolation Schemes.