Adaptive sliding mode observer for sensor fault diagnosis of an industrial gas turbine

• An adaptive sliding mode observer is developed for sensor fault diagnosis of the industrial gas turbine.
• Degradations, faults and noisy outputs are considered without any a priori knowledge about their bounds.
• The finite convergence of the proposed adaptive sliding mode observer is analytically shown.
• The simulation results show the instantaneous detection and fast tracking of the fault profiles for both bias and drift faults.
• The performance of the proposed adaptive sliding mode observer is demonstrated using PROOSIS simulator.

Sensors are one of the crucial components in gas turbines and the failure in sensor measurements can lead to serious problems in maintaining their safety and performance requirements. Our aim in this paper is to develop an adaptive sliding mode observer for sensor fault diagnosis in an industrial gas turbine. The proposed observer has a robustness against gas turbine parameter uncertainties caused by degradations without any priori knowledge about the bounds of faults and parameter uncertainties. The efficiency of the proposed fault diagnosis approach is validated with Matlab/Simulink simulations and the realistic gas turbine data extracted from the PROOSIS software.