Digital Twins-Based Online Monitoring of TFE-731 Turbofan Engine Using Fast Orthogonal Search

Abstract

Due to the complicated structure of the aircraft engine, it is hard to observe the damage and determine its status with the traditional diagnosis method, which sets a fixed threshold for some specific parameters. To accurately reflect the condition of the engine, this article provides a novel fault diagnosis method for the TFE-731 turbofan engine and develops an online diagnosis system. With the combinations of model-based and data-driven approaches, models are constructed to create the so-called digital twins of the engine parameters. For the model-based approach, the physical isentropic compression is used as the basis of the model. For the data-driven approach, the fast orthogonal search is utilized to ensure that the model outputs close to the real engine data. Based on the model prediction-based monitoring strategy, the status of the engine can be identified online. By using the proposed method, an alarm will be triggered once the tested engine outputs cross the boundaries generated by the digital twins. This method can be further applied in quality control to examine the abnormal parts of an aircraft engine. Finally, the digital twins’ online diagnosis system is realized in the practical TFE-731 test facility, verifying the effectiveness of the proposed method.