Knowledge mining for fault diagnosis based on rough sets theory

The fault diagnosis in tribosystem was a difficult problem due to the complex structure of the tribosystem, the nonlinear character of the tribosystem and the presence of multi-excite sources. Usually, one method of fault diagnosis can only inspect one corresponding fault category. In this paper, oil monitoring and vibration monitoring methods were utilized together to diagnose the rolling bearing faults on a homemade bearing bench. Five tests were conducted under different conditions. Oil samples and vibration data were collected regularly and analyzed respectively. Then rough sets theory was introduced into the process of choosing parameters and the knowledge discovery in union diagnosis. Some knowledge was obtained finally.