Sensor Fault Detection, Isolation and Identification Using Multiple Model-based Hybrid Kalman Filter for Gas Turbine Engines (1505.02063v2)

Abstract: In this paper, a novel sensor fault detection, isolation and identification (FDII) strategy is proposed by using the multiple model (MM) approach. The scheme is based on multiple hybrid Kalman filters (HKF) which represents an integration of a nonlinear mathematical model of the system with a number of piecewise linear (PWL) models. The proposed fault detection and isolation (FDI) scheme is capable of detecting and isolating sensor faults during the entire operational regime of the system by interpolating the PWL models using a Bayesian approach. Moreover, the proposed multiple HKF-based FDI scheme is extended to identify the magnitude of a sensor fault by using a modified generalized likelihood ratio (GLR) method which relies on the healthy operational mode of the system. To illustrate the capabilities of our proposed FDII methodology, extensive simulation studies are conducted for a nonlinear gas turbine engine. Various single and concurrent sensor fault scenarios are considered to demonstrate the effectiveness of our proposed on-line hierarchical multiple HKF-based FDII scheme under different flight modes. Finally, our proposed HKF-based FDI approach is compared with various filtering methods such as the linear, extended, unscented and cubature Kalman filters (LKF, EKF, UKF and CKF, respectively) corresponding to both interacting and non-interacting multiple model (MM) based schemes. Our comparative studies confirm the superiority of our proposed HKF method in terms of promptness of the fault detection, lower false alarm rates, as well as robustness with respect to the engine health parameters degradations.

• The paper introduces a multiple model-based hybrid Kalman filter (HKF) integrated with an on-board engine model for sensor fault detection, isolation, and identification across the full operational range of gas turbine engines.
• The HKF method demonstrates robustness against engine health parameter degradation, effectively managing up to 3% estimation errors without false alarms, and offers faster fault detection times compared to other Kalman filter types.
• This robust and efficient HKF approach enhances real-time monitoring and management of gas turbine engines, improving operational efficiency, safety, and overall system reliability by quickly detecting sensor malfunctions.

Sensor Fault Detection, Isolation, and Identification in Gas Turbine Engines Using Hybrid Kalman Filters

This paper introduces a specialized approach for sensor fault detection, isolation, and identification (FDII) in gas turbine engines by utilizing a sophisticated multiple model-based method. The authors propose a hierarchical hybrid Kalman filter (HKF) structure, integrated alongside an on-board engine model (OBEM) and piecewise linear models. This approach enables the effective management of faults over the entire operational range of a gas turbine engine—a domain known for its non-linear characteristics.

The proposed HKF method distinguishes itself by considering parameter degradation due to engine aging and operational wear. By updating the reference baselines for the health parameters, the HKF framework demonstrates resilience against false alarms typically associated with time-varying degradation in health metrics. The method's adaptability in updating the health parameters enables its applicability throughout the gas engine's life cycle.

Key Outcomes and Comparative Analysis

The primary strength of this FDII strategy lies in its promptness in detecting sensor faults and its robustness against health parameter variations. The paper presents simulations covering the full flight profile—climbing, cruise, and landing modes—illustrating the method's capability to handle rapid variations in thrust and ambient conditions. Notably, the approach effectively manages up to 3% health parameter estimation errors without generating false alarms, underscoring its reliability.

In comparative analyses with other filtering techniques—namely linear, extended, unscented, and cubature Kalman filters—the HKF method exhibits faster fault detection times in both interacting and non-interacting multiple model schemes. The HKF, augmented by the OBEM, captures non-linear dynamics more accurately than the MLKF approach, necessitating fewer operational points to maintain estimation accuracy. Furthermore, simulation results indicate that while Extended Kalman Filter (EKF) benefits from updated health parameters for achieving comparable performance, it incurs higher computational burdens compared to HKF.

Implications and Future Work

The implications of this research extend into the field of real-time monitoring and management of gas turbine engines, increasing operational efficiency and safety. The reduction in sensor fault detection time offers significant promise in enhancing system reliability and mitigating risks associated with sensor malfunctions. As the robustness of HKF against health parameters degradation is demonstrated, there exists an opportunity to further refine on-board parameter estimation and control strategies for even larger degradation levels.

Subsequent avenues for investigation might include exploring hybrid sensor-actuator fault schemes, integrating adaptive control mechanisms for fault mitigation, and further reduction of computational overhead without compromising detection speed. Extending the HKF framework to incorporate machine learning methodologies for dynamic modeling and fault prediction could bolster its resilience and accuracy.

In summary, this paper highlights the superiority of the hybrid Kalman filter approach for FDII tasks in complex, non-linear gas turbine systems and paves the way for its adoption in industrial applications demanding high reliability and efficient fault management.