Deep residual shrinkage adaptive network-based cloud-edge collaborative fault diagnosis method for propulsion shafting system

Objectives Aiming at the problems that the fault diagnosis model of propulsion shafting system under variable working conditions has poor generalization and cannot learn autonomously, and that the performance of the model is relatively fixed and cannot be updated online after it is deployed to the edge, a cloud-edge-end collaborative fault diagnosis method based on deep residual shrinkage adaptive network is proposed. Methods First, the historical data of known operating conditions are collected, and a deep residual shrinkage adaptive network model is built in the cloud , through which reinforcement learning algorithms are introduced to enable the model to have the ability of adaptive updating and online learning of the data under the changing working conditions, so as to realize the online updating of the model and the adaptive enhancement of the performance; Then, the model deployment and updating at the edge end is realized by model slice distribution and edge slice aggregation; Finally, real-time fault diagnosis is performed at the edge. Based on the ship propulsion shaft system experimental bench, the effectiveness of the proposed method is verified. Results The results show that the proposed method is able to realize online updating of the model under variable operating conditions, and the updated model has higher fault diagnosis accuracy compared with the un-updated model.?Conclusions The research results can provide a reference for fault diagnosis under variable operating conditions of the propulsion shaft system. Key words：cloud-edge-end collaboration；reinforcement learning；fault diagnosis；propulsion shafting system