A Bank of Kalman Filters for Current Sensors Faults Detection and Isolation of DFIG for Wind Turbine

Abstract

This paper presents a model based Fault diagnosis approach. This approach is based on a bank of Kalman filters. These filters are structured according to the Dedicated Observer Scheme (DOS)., for detecting and isolating multiple and simultaneous current sensors faults of a doubly fed induction generator (DFIG) widely used in variable speed wind turbines. Then., a linear varying parameter model of DFIG is established and the faults diagnosis procedure is implemented. The DFIG model design and the model-based Fault Detection and Isolation (FDI) approach are performed in the Matlab/Simulink environment.