Fault Diagnosis Method For Autonomous Underwater Vehicles Based On Neural Network With Attention Mechanism

Aiming at the autonomous underwater vehicle (AUV), each system module has the characteristics of strong coupling and nonlinearity, which leads to the existence of strong nonlinear relationship between various types of faults and fault features, and the information content of a single signal is limited, which can not meet the diagnosis requirements. For this reason, a fault diagnosis model based on the attention mechanism convolutional neural network (SE-CNN) is proposed. Firstly, the SE attention mechanism module is introduced to fuse the vibration acceleration features in different directions. Second, the convolutional neural network is combined for further feature extraction and fault identification. Finally, experimental validation is carried out by the public dataset of AUV and compared with other state-of-the-art methods. The results show that the SE-CNN model has a good diagnostic effect on the fault data of AUVs and can successfully detect different fault types with high accuracy and stability.