Gas Path On-line Fault Diagnostics Using a Nonlinear Integrated Model for Gas Turbine Engines

Abstract

Gas turbine engine gas path fault diagnosis is closely related technology that assists operators in managing the engine units. However, the performance gradual degradation is inevitable due to the usage, and it result in the model mismatch and then misdiagnosis by the popular model-based approach. In this paper, an on-line integrated architecture based on nonlinear model is developed for gas turbine engine anomaly detection and fault diagnosis over the course of the engine's life. These two engine models have different performance parameter update rate. One is the nonlinear real-time adaptive performance model with the spherical square-root unscented Kalman filter (SSR-UKF) producing performance estimates, and the other is a nonlinear baseline model for the measurement estimates. The fault detection and diagnosis logic is designed to discriminate sensor fault and component fault. This integration architecture is not only aware of long-term engine health degradation but also effective to detect gas path performance anomaly shifts while the engine continues to degrade. Compared to the existing architecture, the proposed approach has its benefit investigated in the experiment and analysis.

Gradient-like minimization methods for aeroengines diagnosis and control

“…Furthermore, all the engine’s component characteristics are strongly non-linear with respect to its operating regime —TIT, EPR, EGT, or whichever variable may be selected to determine the power setting of the turbofan—. Typically, most well-known diagnosis strategies rely on two models: a first, ‘base’ model that computes what would be considered the ‘normal’ operating state of said engine for a given power setting; and then a second, ‘diagnosis’ model —usually making use of linearization or otherwise dimensionality-reduction techniques (Kalman filters come to mind 19 ) to cope with real-time operation— that will compute deviations from the ‘normal’ operating conditions and issue a ‘fault’ signal when the predicted parameters differ more than an established threshold between the ‘base’ model and the ‘diagnosis’ model. This way, the ‘base’ model accounts for the normal aging (slow degradation) process taking place gradually within the engine, 42 while the ‘diagnosis’ model would detect faults suddenly occurring in the engine components and issue the corresponding signal.…”

“…Working with measurements from installed sensors and with some strong simplifying assumptions in their models to alleviate the computational load (no use of the turbomachinery component maps is made of, for instance), which limit the accuracy of the predictions to some extent, the researchers make use of several Kalman filter techniques to achieve real-time update of the fault thresholds and to diagnose all sensors installed within the engine. A similar approach to diagnose aeroengine components was also presented by Lu et al 19 In addition, Li et al 17 presented a new version of the GPA method for the calculation of the performance of an engine, to predict and control its behavior. The aim is to be able to fine-tune an engine fleet model to accurately represent each individual engine within the fleet, by means of model adaption.…”

Neural-Learning-Based Telerobot Control With Guaranteed Performance

“…In order to smoothly switch in between the obstacle avoidance and the restoration, we employ a weighted sum of (17) and (18), and integrate it into the design of the desired joint velocities aṡ…”

Fault diagnosis based on grey relational analysis and synergetic pattern recognition for aero-engine gas-path systems

“…According to the recognition process of synergetic fault patterns, these types of faults are difficult to be distinguished from each other, so they must be discriminated further. A7 is taken as an example, and the prototype vector v k is normalized and zero-averaged The initial order parameter is evolved according to equations (3) and 4, and its evolving process is shown in Figure 2. Figure 2 shows that the order parameter value for fault type A7 is 1.…”

“…Finding a fault diagnosis method, which can accurately determine fault types, is worth studying. At present, there are many fault diagnosis methods for gas-path systems, including the analytic model-driven method, [3][4][5] knowledgedriven method, 6,7 and data-driven method. 8,9 Due to the complexity of gas-path systems, the model-driven method has poor fault diagnosis accuracy because it is difficult to build an accurate analytic model for a gaspath system.…”