Intelligent Fault Diagnosis of Liquid Rocket Engine via Interpretable LSTM with Multisensory Data

• Published in: Sensors (Basel, Switzerland) Vol. 23; no. 12; p. 5636
• Main Authors: Zhang, Xiaoguang, Hua, Xuanhao, Zhu, Junjie, Ma, Meng
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 16.06.2023; MDPI
• Subjects: Algorithms; Analysis; Artificial intelligence; bidirectional LSTM; data fusion; Deep learning; Efficiency; Fault diagnosis; fault simulation; Hydrogen; interpretable; Machine learning; Mathematical models; Neural networks; Power; Rockets; Simulation; Turbines; data fusion; interpretable; fault simulation; bidirectional LSTM

Fault diagnosis is essential for high energy systems such as liquid rocket engines (LREs) due to harsh thermal and mechanical working environment. In this study, a novel method based on one-dimension Convolutional Neural Network (1D-CNN) and interpretable bidirectional Long Short-term Memory (LSTM)...