Fault Detection and Isolation for Wind Turbine Systems Based on Proportional Multi-Integral Observer (PMIO)

Abstract

This paper addresses Fault Detection and Isolation (FDI) for wind turbines based on a Proportional 
Multi-Integral Observer (PMIO). A wind turbine model is linearized using the Takagi-Sugeno (TS) approach based on 
Lyapunov stability theory and LMI condition, then the PMI observer is considered for use with the TS fuzzy model to 
estimate and isolate both actuator and sensor faults with the introduction of a centered noise. The kth derivatives of 
the actuators and sensor faults are not equal to zero but are rather bounded norms. However, based on Lyapunov 
stability theory and L2 performance analysis, design conditions are established through LMIs formulations. Simulation 
results show that our proposal outperforms some existing approaches.