In a groundbreaking development for the aviation industry, researchers have introduced an intelligent fault diagnosis framework specifically designed for general aviation aircraft. This innovative method, detailed in a recent paper on arXiv, addresses the pressing challenges of scarce real fault data, diverse fault types, and weak fault signatures that have long plagued aircraft maintenance and safety protocols.

The proposed framework leverages a multi-fidelity digital twin approach, integrating four key modules: high-fidelity flight dynamics simulation, FMEA-driven fault injection, multi-fidelity residual feature extraction, and a large language model (LLM)-enhanced report generation system. At its core, the digital twin is built using the JSBSim six-degree-of-freedom (6-DoF) flight dynamics engine, which generates comprehensive engine health monitoring data through advanced sensor synthesis equations.

One of the standout features of this framework is its three-layer fault injection engine, which employs failure mode and effects analysis (FMEA) to model the physical causal propagation of 19 different engine fault types. Additionally, a multi-fidelity residual computation framework utilizes paired-mirror residuals and GRU surrogate prediction residuals to enhance diagnostic accuracy. Notably, the high-fidelity path captures clean fault deviation signals, while the low-fidelity path enables real-time residual computation.

With a 1D-CNN classifier performing end-to-end diagnosis across 20 fault classes, the system demonstrates impressive results, achieving a Macro-F1 score of 96.2%. Furthermore, the GRU surrogate scheme boasts a remarkable 4.3x inference acceleration with minimal performance trade-offs.

This pioneering research not only sets a new standard for aircraft fault diagnosis but also emphasizes the importance of residual feature quality, establishing a design principle that prioritizes diagnostic performance. As the aviation industry continues to embrace AI technologies, this advancement promises to enhance safety and reliability in general aviation.