Machines which are developed today are highly automated due to increased use of mechatronic systems. Fault detection and isolation are important features to ensure their reliable operation. This research work aims to achieve an integrated control and fault detection functionality with minimum number of components. Piezomaterials can be used both as sensors and actuators due to their inherent coupling between electrical and mechanical properties. In this thesis, piezoelectric actuators which are being researched for active vibration control on flexible rotors, are also used as sensors. Since the actuators are mounted in rotor bearings, their displacement represent the bearing deflection directly. The virtual sensor signals reconstructed from actuators are utilised for model-based fault diagnosis in rotor systems, with focus on unbalances in rotors. The robustness of the estimates against signal outliers is improved by method of M-estimators. Identifying the fault location is also explored using the hypothesis of localization of faults. The systems under investigation are a numerical aircraft engine model and its scaled representation in a rotor test bench.
