Remaining useful life prediction method for rotating machinery based on multi-stage information fusion

Deep learning algorithms have been widely used in the field of remaining useful life (RUL) prediction for rotating machinery in recent years due to their powerful feature extraction and self-learning capabilities. However, the increasingly complex industrial environment puts forward higher requirements on the RUL prediction accuracy of equipment. For this reason, this study improves the algorithm of Transformer class based on the global-local information fusion method to achieve the RUL prediction of bearings. Using Transformer model for local life prediction, frequency enhanced decomposed Transformer (FEDformer) model for global life prediction and particle swarm optimization (PSO) algorithm to optimize the weighting coefficients, a multi-stage information fusion RUL method, named multi-stage fusion Transformer (MSFformer), is designed, which realises the optimal combination of global and local information and improves the prediction accuracy. The effectiveness of the proposed method is verified on the XJTU-SY dataset.