Model-Based Fault Diagnosis Techniques

Steven X. Ding began investigating issues concerned with model-based fault detection and identification in 1986 during his PhD. Following a three-year stay in industry, in which he gained experience with the application of FDI techniques in real technical processes, he returned to academia and during the last 17 years has become a university professor and institute head. He has been involved with numerous national and international research grants and industrial projects in the development of advanced FDI methods and their application in different sectors of industry. He teaches fault diagnosis and fault-tolerant systems to Masters students and advanced FDI methods to PhD students at the University of Duisburg-Essen. He has also guest-lectured and taught courses on these subjects at other universities and research institutes. Professor Ding has published more than 80 journal and 130 conference papers associated with the field of FDI.

• 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.