Fault Detection and Prognosis of Aeroengine Fuel Pipelines Based on Deep Learning

The reliability of aeroengine fuel pipeline systems is directly related to flight safety. Traditional fault detection approaches suffer from low accuracy and poor real-time performance. This paper presents a deep learning-based methodology for fault detection and prognosis of aeroengine fuel pipelines. A comprehensive monitoring system is constructed using multi-sensor data fusion technology, a fault identification model based on LSTM-CNN hybrid neural networks is designed, and a fault prediction algorithm based on time series analysis is established. Simulation experiments demonstrate that the proposed method achieves fault detection accuracy of 96.847 percent and fault prediction precision of 94.235 percent, representing improvements of 12.584 percent and 15.792 percent respectively compared to traditional methods. Research results indicate that the deep learning-based approach can effectively enhance the accuracy and real-time performance of aeroengine fuel pipeline fault detection and prognosis, providing significant technical support for aeroengine health management.