Fuzzy observer-based fault detection for wind turbines using the finite-frequency approach

Abstract

ABSTRACT This research investigates the synthesis of a new Finite Frequency Observer (FFO) to detect faults affecting wind turbines. First, the wind turbine nonlinear behaviour is approximated using Takagi-Sugeno (T-S) fuzzy modelling technique for observer design purposes. Then, through the H-/H∞ optimisation technique with Finite Frequency (FF) specification, the FFO design conditions are formulated using the Lyapunov method as a set of LMIs. Finally, to assess the efficiency of the FFO-based detection approach, simulations results are applied to a dynamic 4.8 MW WT model, where an Unknown Input Observer method is used for comparison.