Design of Turbine Engine Robust Fault Detection with LPV Model

This paper deals with a problem of turbine engine fault detection in dynamic process by using LPV mathematical model to exactly describe turbine engine dynamic characteristics, eliminate the effects of modeling error. Design robust fault detection filter (FDF) with eigenstructure assignment and realizes disturbance de-coupled with model uncertainty. The simulation results of turbine engine control system sensor fault show that this method has good performance in focusing discrimination in fault signal with modeling error, enhancing the robustness to unknown input, detecting accuracy is high and satisfiying real-time requirement.