Dynamic Decision Support and Automated Fault Accommodation for Jet Engines

The development of a dynamic decision support (D²S) system for real-time assessments of system health and fault contingency planning for jet engine is presented. The goal of the proposed D²S system is to increase engine operation autonomy level by bridging the gap between onboard prognosis & health management (PHM) system and baseline control systems. The D²S system consists of two major functions: real-time system health identification (RT-SHI), and automated contingency management (ACM). The RT-SHI modules enhance on-board PHM functions with a dynamic system identification algorithm that is capable of detecting and isolating faults/failures with a continuously updated dynamic model. Particularly, a real-time strong tracking system identification module that is robust to low-excitation conditions was developed to track time-varying parameters and sudden changes in plant dynamics. In addition, a realtime, self-tuning Kalman filter and a probabilistic neural networks based fault classifier are combined to provide accurate health estimation. Based on the inferred health condition, mission requirements and flight regime information, the on-board ACM module automatically makes decisions regarding control reconfiguration and change of control strategies. The presented D²S architecture and modules have been applied to a generic turbofan engine model. Simulation results are presented to illustrate the effectiveness of the approaches.