Fault Diagnosis of Rolling Bearing Based on Rough Set and Neural Network

Artificial neural network was one of the most important methods in intelligent fault diagnosis because it has the performance of nonlinear pattern classification and the capacity of self-learning and self-organization, but it cannot judge redundancy and usefulness of information. Rough set can reduce the knowledge of information system and dislodge redundant information. In this paper, fault data of rolling bearing was reduced by the greedy algorithm of rough set. Training data and test data of BP neural network had been reduced by rough set. By comparison of two test result about simply data and original data, it was indicated that resolving power was unchanged and database was simply.