Research on Fault Diagnosis Technology of Inertial Navigation Systems on Aircraft

Inertial navigation systems play a crucial role in the aviation industry, and their safe operation is of paramount importance to the entire flight mission. The research on fault diagnosis of inertial navigation systems is becoming increasingly significant. This paper conducts an in-depth study on the fault diagnosis of strapdown inertial navigation systems in fault-tolerant integrated navigation systems. The paper thoroughly analyzes the fault modes and effects of strapdown inertial navigation systems, determines the system's composition structure, identifies possible fault causes and mitigation solutions, and proposes a fault diagnosis process. It conducts an in-depth analysis of the fault manifestations of relevant components and determines the forms of fault manifestations. Based on the fault manifestations, data collection and preprocessing are performed using simulation equipment and relevant datasets to obtain operational data of the inertial navigation system under various working conditions. This paper builds a model for diagnosing component faults in inertial navigation systems using deep learning methods. In terms of selecting the deep learning model, this paper constructs a Transformer model for fault diagnosis classification. After building the model, through validation and analysis of experimental data, this paper demonstrates the effectiveness and feasibility of the model used in fault diagnosis of strapdown inertial navigation systems, proving that the constructed model can diagnose faults in inertial navigation systems. Finally, the study is summarized, and future research directions in this field are discussed.