Intelligent fault diagnosis method of rolling bearing based on multi-source domain fast adversarial network

Abstract

Fault diagnosis of rolling bearings are among the most crucial links in the prognostic and health management (PHM) of bearings. To solve the problem that single-source domain transfer learning can not adapt well to the target domain, a transfer diagnosis method based on Multi-source Domain Fast Adversarial Network (MSDFAN) is proposed. Firstly, the signals from all domains are input into a common subnetwork of fast neural networks to reduce the complexity and network running time of neural networks. Secondly, several adversarial networks are constructed as domain specific feature extractors and then use Higher-order Moment Matching (HoMM) to reduce distribution differences between A and B domains. The two experimental cases of rolling bearing supports the effectiveness and superiority of the proposed MSDFAN.

TSMDA: intelligent fault diagnosis of rolling bearing with two stage multi-source domain adaptation

“…However, the adaptation capability of general methods may be insufficient due to the limitation of knowledge transfer from the single-source. Therefore, multisource data from various scenarios can be collected to provide richer generalized knowledge [26][27][28][29]. Li and Yu [30] proposed a TL model based on feature-level and class-level adaptations for bearing fault diagnosis under different working conditions.…”