A hybrid fine-tuned VMD and CNN scheme for untrained compound fault diagnosis of rotating machinery with unequal-severity faults

作者：

Highlights：

• Capability of identifying minor faults overshadowed by more severe ones.

• CNN training process only using the healthy and single failures dataset.

• Using the compound failure data only in the CNN test stage.

• VMD parameter optimization by fine-tuned VMD method based on adaptive indices.

摘要

•Capability of identifying minor faults overshadowed by more severe ones.•CNN training process only using the healthy and single failures dataset.•Using the compound failure data only in the CNN test stage.•VMD parameter optimization by fine-tuned VMD method based on adaptive indices.

论文关键词：Fault diagnosis,Rotating machinery,Compound fault,Variational mode decomposition,Convolutional neural network

论文评审过程：Received 6 April 2020, Revised 16 August 2020, Accepted 2 October 2020, Available online 22 October 2020, Version of Record 10 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114094