Three-stage algorithm of estimation and fault diagnose for closed-loop gas turbine engine systems with unknown time delay

This paper proposes a three-stage estimation algorithm for identification and fault diagnose of closed-loop linear systems of gas turbine engines with unknown time delay, which can be used for estimating parameters of systems and faults. By means of calculating the ratio of data determinant, the orders of time delay and the system are simultaneously estimated in the first stage. Then a two-stage least square identification method is applied to estimating the parameters and fault of an auxiliary CARMA model with time delay has been calculated. The proposed algorithm has lower computational cost and is effective for gas turbine engine systems with unknown time delay. Finally, a numerical example of biaxial gas turbine engine system is given to shown the effectiveness of the proposed method.