Adaptive Observer Based Fault Detection and Isolation for Wind Turbines

Fault diagnosis strategies are in high demand to reduce the down-times and improve the reliability of wind turbines. An adaptive observer method is proposed for fault detection and isolation (FDI) of the wind turbine benchmark model. Firstly, the state spaces of wind turbine subsystems are remodeled according to the output values of sensors. Corresponding adaptive observers are then designed to estimate subsystem states and detect faults simultaneously. To speed up the fault diagnosis process and to improve the precision, the fast adaptive fault estimation (FAFE) algorithm is adopted in the adaptive observers. Besides, a reduced-order model is introduced for the drive train subsystem to eliminate the disturbance of unknown aerodynamics. Simulation results illustrate that all the eight fault types defined in the benchmark model of wind turbines can be successfully detected.