Unanticipated Fault Detection and Isolation of Telescope Drive System Based on Luenberger Observer and Axes Transformation

This paper puts forward a new strategy current sensor unanticipated faults detection and isolation (FDI) for permanent magnet synchronous motor (PMSM) in a telescope drive system. This approach uses axes transformation, PMSM model and Luenberger observer to generate residuals, and the influence of unanticipated faults (UFs) in different phases on the current components is analyzed. The sensor UFs detection is performed by processing residuals obtained from the observer. In addition, based on the information provided by fault detection, an innovative logic judgment algorithm is devised to realize fault isolation. The proposed method can discriminate between single UF of different types and multiple simultaneous UFs of various categories as well as faulty current sensors. Extensive simulation experiments prove that the designed logic judgment algorithm is effective, and the FDI can be implemented successfully.