Synthesizing Rolling Bearing Fault Samples in New Conditions: A framework based on a modified CGAN. (arXiv:2206.12076v2 [cs.LG] UPDATED)

Bearings are one of the vital components of rotating machines that are prone
to unexpected faults. Therefore, bearing fault diagnosis and condition
monitoring is essential for reducing operational costs and downtime in numerous
industries. In various production conditions, bearings can be operated under a
range of loads and speeds, which causes different vibration patterns associated
with each fault type. Normal data is ample as systems usually work in desired
conditions. On the other hand, fault data is rare, and in …

arxiv framework lg