Fault detection and isolation using Kalman filter bank for a wind turbine generator

In this paper, we will tackle the problem of current sensor Fault Detection and Isolation (FDI) of a doubly feed induction generator in wind turbine. First of all, the use of a Kalman filter to detect sensor fault will be illustrated. Then, the detection and the isolation of multiple and simultaneous sensor faults will be treated using a Kalman filter bank in a Generalized Observer Scheme and a Dedicated Observer Scheme respectively. The fault assessment will be performed by the Page-Hinkley's test. The processes of the FDI will be validated on a real-time test bench.