AUV thruster fault diagnosis based on comprehensive model

Fault diagnosis model is an effective method to ensure the long-term reliable operation of AUV (Autonomous Underwater Vehicle, AUV) but currently there is still a lack of comprehensive and practical system methods. Therefore, a comprehensive AUV fault diagnosis method was adopted, and the method was illustrated and validated using the AUV thruster system fault diagnosis as a case study. This method closely links the control module, hardware module, communication module and navigation module through knowledge transfer. Comprehensive methods such as stochastic resonance, neural network, principal component analysis, bayesian network, and markov chains for fault diagnosis. The obtained fault model provides possibilities for subsequent online optimization, which also promotes the development of large-scale fault diagnosis model.