Adaptive mask flow adversarial network for single-source domain generalization fault diagnosis of train bearings

2026-08-12

PRE

王

王俊 (Jun Wang)

Bochao Yu

He Ren

Zhongkui Zhu

Yifan Huangfu *

黄伟国 (Weiguo Huang)

DOI：10.1016/j.isatra.2026.08.010

Original

Original request for help

Save

Abstract

En 中文

• Single-source train bearing state information is used for generalization diagnosis. • Simulation of domain shifts is conducted in feature space instead of data space. • Adaptive mask mechanism is designed to determine the domain-sensitive features. • Flow-based feature perturbation strategy enables unseen domain generalization. • Superior diagnosis generalization ability to unseen working conditions is verified.

Keywords:

Single-source domain generalization

train bearing fault diagnosis

flow model

adaptive mask

generative adversarial network

Adaptive mask flow adversarial network for single-source domain generalization fault diagnosis of train bearings

Abstract

Journal

Organization

Cited Papers

Citing Papers