Application of transient acoustic signal and rough sets in compound-fault diagnosis of bearing

Real vehicle tests for equipment cannot get steady signal, and the traditional vibration test method cannot realize noncontact, without disassembly and spoil the disadvantages of the on-line detection, using the transient acoustic signal is proposed for bearing fault diagnosis. Rough set theory is a new mathematical tool to deal with vagueness and uncertainty knowledge, attribute reduction is an important part of the study of rough sets theory, and calculates the best reduction is a NP hard problem. In order to be able to effectively obtain information system of the reduction, a new reduction algorithm is proposed. The algorithm with the worst Fisher criterion function as heuristic information to improve the efficiency of search, the Max-min ant system (MMAS) conditional attribute reduction, the use of rough set theory to fault diagnosis decision table reduction, form a clear, concise fault diagnosis rules, effectively improve the bearing fault diagnosis accuracy and efficiency.