Online Identification Technology for Aircraft Power Failure Based on LSTM Neural Network and Inertial Information

This paper discusses a typical thrust fault online identification method for aircraft during flight, based on the aircraft motion information of the control system and the Long Short-Term Memory (LSTM) neural network. Aimed at the typical power failure mode of liquid main engine thrust decline, the paper studies the technology of typical power failure identification, and completes the demonstration and verification of key technologies through flight tests using an aircraft. On the basis of limited flight computing power, the paper completes the research on fault identification methods based on neural network methods. By using the inertial guidance information of the flight, it identifies the problem of flight failure caused by non-fatal power failures. It further enhances the autonomous adaptability and intelligent level of the aircraft.