A novel approach to aeroengine performance diagnosis based on physical model coupling data-driven using low-rank multimodal fusion method

Aeroengine health assessment is crucial to ensure flight safety and reliability. Traditionally, it involves diagnosing the performance of the aeroengine gas path. However, due to the complexity of operating conditions, non-linear performance, and the coupling of gas path performance fault characteristics, determining the aeroengine health condition directly from engine monitoring information is challenging, especially with insufficient sensor data. To tackle these challenges, a novel digital twin method for aeroengine performance diagnosis was proposed. This method utilizes data-driven and performance models, employing low-rank multimodal fusion. By digitizing the physical system or process through mathematical models and simulation technology, this approach offers distinct advantages compared to previous methods based solely on models or data. At the aeroengine component level, an adaptive model was employed, and the data-driven model was constructed using flight data. Support vector machines were utilized for gas path fault classification. The engine digital twin was created through low-order multimodal fusion. The results demonstrate that the proposed method achieves excellent diagnostic accuracy under both steady and transient conditions. It can be utilized to enhance engine performance monitoring and evaluation, as well as improve the reliability, availability, and efficiency of the engine.